Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
ISSN 1937 - 1055
VOLUME 4, 2019
INTERNATIONAL JOURNAL OF
MATHEMATICAL COMBINATORICS
EDITED BY
THE MADIS OF CHINESE ACADEMY OF SCIENCES AND
ACADEMY OF MATHEMATICAL COMBINATORICS & APPLICATIONS, USA
December, 2019
Vol.4, 2019 ISSN 1937-1055
International Journal of
Mathematical Combinatorics
(www.mathcombin.com)
Edited By
The Madis of Chinese Academy of Sciences and
Academy of Mathematical Combinatorics & Applications, USA
December, 2019
Aims and Scope: The mathematical combinatorics is a subject that applying combinatorial
notion to all mathematics and all sciences for understanding the reality of things in the universe,
motivated by CC Conjecture of Dr.Linfan MAO on mathematical sciences. The International
J.Mathematical Combinatorics (ISSN 1937-1055) is a fully refereed international journal,
sponsored by the MADIS of Chinese Academy of Sciences and published in USA quarterly,
which publishes original research papers and survey articles in all aspects of mathematical
combinatorics, Smarandache multi-spaces, Smarandache geometries, non-Euclidean geometry,
topology and their applications to other sciences. Topics in detail to be covered are:
Mathematical combinatorics;
Smarandache multi-spaces and Smarandache geometries with applications to other sciences;
Topological graphs; Algebraic graphs; Random graphs; Combinatorial maps; Graph and
map enumeration; Combinatorial designs; Combinatorial enumeration;
Differential Geometry; Geometry on manifolds; Low Dimensional Topology; Differential
Topology; Topology of Manifolds;
Geometrical aspects of Mathematical Physics and Relations with Manifold Topology;
Mathematical theory on gravitational fields and parallel universes;
Applications of Combinatorics to mathematics and theoretical physics.
Generally, papers on applications of combinatorics to other mathematics and other sciences
are welcome by this journal.
It is also available from the below international databases:
Serials Group/Editorial Department of EBSCO Publishing
10 Estes St. Ipswich, MA 01938-2106, USA
Tel.: (978) 356-6500, Ext. 2262 Fax: (978) 356-9371
http://www.ebsco.com/home/printsubs/priceproj.asp
and
Gale Directory of Publications and Broadcast Media, Gale, a part of Cengage Learning
27500 Drake Rd. Farmington Hills, MI 48331-3535, USA
Tel.: (248) 699-4253, ext. 1326; 1-800-347-GALE Fax: (248) 699-8075
http://www.gale.com
Indexing and Reviews: Mathematical Reviews (USA), Zentralblatt Math (Germany), Refer-
ativnyi Zhurnal (Russia), Mathematika (Russia), Directory of Open Access (DoAJ), EBSCO
(USA), International Scientific Indexing (ISI, impact factor 1.972), Institute for Scientific In-
formation (PA, USA), Library of Congress Subject Headings (USA).
Subscription A subscription can be ordered by an email directly to
Linfan Mao
The Editor-in-Chief of International Journal of Mathematical CombinatoricsChinese Academy of Mathematics and System Science Beijing, 100190, P.R.China, and also thePresident of Academy of Mathematical Combinatorics & Applications (AMCA), Colorado, USA
Email: [email protected]
Price: US$48.00
Editorial Board (4th)
Editor-in-Chief
Linfan MAO
Chinese Academy of Mathematics and System
Science, P.R.China
and
Academy of Mathematical Combinatorics &
Applications, Colorado, USA
Email: [email protected]
Deputy Editor-in-Chief
Guohua Song
Beijing University of Civil Engineering and
Architecture, P.R.China
Email: [email protected]
Editors
Arindam Bhattacharyya
Jadavpur University, India
Email: [email protected]
Said Broumi
Hassan II University Mohammedia
Hay El Baraka Ben M’sik Casablanca
B.P.7951 Morocco
Junliang Cai
Beijing Normal University, P.R.China
Email: [email protected]
Yanxun Chang
Beijing Jiaotong University, P.R.China
Email: [email protected]
Jingan Cui
Beijing University of Civil Engineering and
Architecture, P.R.China
Email: [email protected]
Shaofei Du
Capital Normal University, P.R.China
Email: [email protected]
Xiaodong Hu
Chinese Academy of Mathematics and System
Science, P.R.China
Email: [email protected]
Yuanqiu Huang
Hunan Normal University, P.R.China
Email: [email protected]
H.Iseri
Mansfield University, USA
Email: [email protected]
Xueliang Li
Nankai University, P.R.China
Email: [email protected]
Guodong Liu
Huizhou University
Email: [email protected]
W.B.Vasantha Kandasamy
Indian Institute of Technology, India
Email: [email protected]
Ion Patrascu
Fratii Buzesti National College
Craiova Romania
Han Ren
East China Normal University, P.R.China
Email: [email protected]
Ovidiu-Ilie Sandru
Politechnica University of Bucharest
Romania
ii International Journal of Mathematical Combinatorics
Mingyao Xu
Peking University, P.R.China
Email: [email protected]
Guiying Yan
Chinese Academy of Mathematics and System
Science, P.R.China
Email: [email protected]
Y. Zhang
Department of Computer Science
Georgia State University, Atlanta, USA
Famous Words:
If you want to know everything all at once, you will know nothing.
By Ivan Pavlov, a former Soviet physiologist, psychologist.
International J.Math. Combin. Vol.4(2019), 1-18
Graphs, Networks and Natural Reality
– from Intuitive Abstracting to Theory
Linfan MAO
1. Chinese Academy of Mathematics and System Science, Beijing 100190, P.R.China
2. Academy of Mathematical Combinatorics & Applications (AMCA), Colorado, USA
E-mail: [email protected]
Abstract: In the view of modern science, a matter is nothing else but a complex network−→G , i.e., the reality of matter is characterized by complex network. However, there are no
such a mathematical theory on complex network unless local and statistical results. Could we
establish such a mathematics on complex network? The answer is affirmative, i.e., mathemat-
ical combinatorics or mathematics over topological graphs. Then, what is a graph? How does
it appears in the universe? And what is its role for understanding of the reality of matters?
The main purpose of this paper is to survey the progressing process and explains the notion
from graphs to complex network and then, abstracts mathematical elements for understand-
ing reality of matters. For example, L.Euler’s solving on the problem of Konigsberg seven
bridges resulted in graph theory and embedding graphs in compact n-manifold, particularly,
compact 2-manifold or surface with combinatorial maps and then, complex networks with
reality of matters. We introduce 2 kinds of mathematical elements respectively on living
body or non-living body for self-adaptive systems in the universe, i.e., continuity flow and
harmonic flow−→GL which are essentially elements in Banach space over graphs with operator
actions on ends of edges in graph−→G . We explain how to establish mathematics on the 2
kinds of elements, i.e., vectors underling a combinatorial structure−→G by generalize a few
well-known theorems on Banach or Hilbert space and contribute mathematics on complex
networks. All of these imply that graphs expand the mathematical field, establish the foun-
dation on holding on the nature and networks are closer more to the real but without a
systematic theory. However, its generalization enables one to establish mathematics over
graphs, i.e., mathematical combinatorics on reality of matters in the universe.
Key Words: Graph, 2-cell embedding of graph, combinatorial map, complex network,
reality, mathematical element, Smarandache multispace, mathematical combinatorics.
AMS(2010): 00A69,05C21,05C25,05C30 05C82, 15A03,57M20
§1. Introduction
What is the role of mathematics to natural reality? Certainly, as the science of quantity,
mathematics is the main tool for humans understanding matters, both for the macro and the
micro in the universe. Generally, it builds a model and characterizes the behavior of a matter
1Received August 25, 2019, Accepted November 20, 2019.
2 Linfan MAO
for holding on reality and then, establishes a theory, such as those shown in Fig.1.�� ���� ���� ���� ��- - -Observation& Experiment
Induction& Deduction
Hypothesis& Testing Theory
Fig.1
This scientific method on matters in the universe is completely reflected in the solving
process of L.Euler on the problem of Konigsberg seven bridges. Geographically, the city of
Konigsberg is located on both sides of Pregel River, including two large islands which were
connected to each other and the mainland by seven bridges, such as those shown in Fig.2.
The residents of Konigsberg usually wished to pass through each bridge once without repeat,
initialing at point of the mainland or islands.
Fig.2
However, no one traveled in such a way once. Then, a resident should how to travel for such a
walk? L.Euler solved this problem, and answered it had no solution in 1736. How did he do it?
Let A,B,C,D be the two sides and islands. Then, he abstracted this problem on (a) equivalent
to finding a traveling passing through each lines on (b) without repeating.
(a) (b)
Fig.3
Clearly, such a traveling must be with the same in and out times at each point A,B,C or D.
But, (b) is not fitted with such conditions. So, there are no such a traveling in the problem on
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 3
Konigsberg seven bridges.
Euler’s solving method on the problem of Konigsberg seven bridges finally resulted graph
theory into beings today. A graph G is an ordered 3-tuple (V,E; I), where V,E are finite sets,
V 6= ∅ and I : E → V × V . Call V the vertex set and E the edge set of G, denoted by V (G)
and E(G), respectively. For example, two graphs K(3, 3) and K5 are shown in Fig.4.r r rr r r r r rr rK(3, 3) K5
Fig.4
Usually, if (u, v) = (v, u) for ∀u, v ∈ V (G), then G is called a graph. Otherwise, it is called
a directed graph with an orientation u→ v on each edge (u, v), denoted by−→G .
Let G1 = (V1, E1, I1), G2 = (V2, E2, I2) be 2 graphs. If there exists a 1 − 1 mapping
φ : V1 → V2 and φ : E1 → E2 such that φI1(e) = I2φ(e) for ∀e ∈ E1 with the convention that
φ(u, v) = (φ(u), φ(v)), then we say that G1 is isomorphic to G2, denoted by G1∼= G2 and φ
an isomorphism between G1 and G2. Clearly, all automorphisms φ : V (G) → V (G) of graph G
form a group under the composition operation, and denoted by AutG the automorphism group
of graph G. A few automorphism groups of well-known graphs are listed in Table 1.
G AutG order
Pn Z2 2
Cn Dn 2n
Kn Sn n!
Km,n(m 6= n) Sm × Sn m!n!
Kn,n S2[Sn] 2n!2
Table 1
Certainly, an edge e = uv ∈ E(G) can be divided into two semi-arcs eu, ev such as those
shown in Fig.5.u uu v
- �u uu v
- eu eve Divided into
Fig.5
Similarly, two semi-arcs eu, fv are called v-incident or e-incident if u = v or e = f . Denote
all semi-arcs of a graph G by X 12(G). A 1−1 mapping ξ on X 1
2(G) such that ∀eu, fv ∈ X 1
2(G),
ξ(eu) and ξ(fv) are v−incident or e−incident if eu and fv are v−incident or e−incident, is
4 Linfan MAO
called a semi-arc automorphism of the graph G. Clearly, all semi-arc automorphisms of a graph
also form a group, denoted by Aut 12G.
Certainly, graph theory studies properties of graphs. A property is nothing else but a
family of graph, i.e., P = {G1, G2, · · · , Gn, · · · } but closed under isomorphisms φ of graphs,
i.e., Gφ ∈ P if G ∈ P. For example, hamiltonian graphs, Euler graphs and also interesting
parameters, such as those of connectivity, independent number, covering number, girth, level
number, · · · of a graph.
The main purpose of this paper is to survey the progressing process and explains the notion
from graphs to complex network and then, abstracts mathematical elements for understanding
reality of matters. For example, L.Euler’s solving on the problem of Konigsberg seven bridges
resulted in graph theory and embedding graphs in compact n-manifold, particularly, compact
2-manifold or surface with combinatorial maps and then, complex networks with reality of
matters. We introduce 2 kinds of mathematical elements respectively on living or non-living
body in the universe, i.e., continuity and harmonic flows−→GL which are essentially elements
in Banach space over graphs with operator actions on ends of edges in graph−→G . We explain
how to establish mathematics on the 2 kinds of elements, i.e., vectors underling a combinatorial
structure−→G by generalize a few well-known theorems on Banach or Hilbert space and contribute
a mathematics on complex networks.
For terminologies and notations not mentioned here, we follow references [1],[2] and [4] for
graphs, [3] for complex network, [6] for automorphisms of graph, [24] for algebraic topology,
[25] for elementary particles and [6],[26] for Smarandache systems and multispaces.
§2. Embedding Graphs on Surfaces
2.1 Surface
A surface is a 2-dimensional compact manifold without boundary. For example, a few surfaces
are shown in Fig.6.
Sphere Torus Klein bottle
Fig.6
Clearly, the intuition imagination is difficult for determining surface of higher genus. How-
ever, T.Rado showed the following result, which is the fundamental of combinatorial topology,
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 5
ro topological graphs on surfaces.
Theorem 2.1(T.Rado 1925,[24]) For any compact surface S, there exist a triangulation {Ti, i ≥1} on S.
T.Rado’s result on triangulation of surface enables one to present a surface by listing every
triangle with each side a label and a direction, i.e., the polygon representation. Then, the surface
is assembled by identifying the two sides with the same label and direction. This way results
in a polygon representation on a surface finally. For examples, the polygon representations on
surfaces in Fig.6 are shown in Fig.7.
����- 6 6 6-
---6 ?a a
B
A
a a
b
b
b
b
a a
Sphere Torus Klein bottle
Fig.7
We know the classification theorem of surfaces following.
Theorem 2.2([24]) Any connected compact surface S is either homeomorphic to a sphere, or to
a connected sum of tori, or to a connected sum of projective planes, i.e., its surface presentation
S is elementary equivalent to one of the standard surface presentations following:
(1) The sphere S2 =⟨a|aa−1
⟩;
(2) The connected sum of p tori
T 2#T 2# · · ·#T 2
︸ ︷︷ ︸p
=
⟨ai, bi, 1 ≤ i ≤ p |
p∏
i=1
aibia−1i b−1
i
⟩;
(3) The connected sum of q projective planes
P 2#P 2 · · ·#P 2
︸ ︷︷ ︸q
=
⟨ai, 1 ≤ i ≤ q |
q∏
i=1
ai
⟩.
A combinatorial proof on Theorem 2.2 can be found in [6]. By definition, the Euler
characteristic of S is
χ(S) = |V (S)| − |E(S)| + |F (S)|,
where V (S), E(S) and F (S) are respective the set of vertex set, edge set and face set of the
polygon representation of surface S. Then, we know the next result.
6 Linfan MAO
Theorem 2.3([24]) Let S be a connected compact surface with a presentation S. Then
χ(S) =
2, if S ∼El S2,
2 − 2p, if S ∼El T2#T 2# · · ·#T 2
︸ ︷︷ ︸p
,
2 − q, if S ∼El P2#P 2# · · ·#P 2
︸ ︷︷ ︸q
.
Theorem 2.3 enables one to define the genus of orientable or non-orientable surface S by
numbers p and q, respectively, and the genus of sphere is defined to be 0.
2.2 Embedding Graph
A graph G is said to be embeddable into a topological space T if there is a 1 − 1 continuous
mapping φ : G → T with φ(p) 6= φ(q) if p, q are different points on graph G. Particularly, if
T = R2 is a Euclidean plane, we say that G is a planar graph.
A most interesting case on the embedding problem of graphs is the case of surface, which is
essentially to search the polyhedral structures on surfaces. Clearly, many results on embedding
graphs is on these surfaces with small genus. For example, embedding results on p = 0 the
sphere, p = 1 the torus, · · · of orientable surfaces, or on q = 1 the projective plane, q = 2
the Kein bottle, · · · of non-orientable surfaces. The most simple case is embedding graphs on
sphere which is equivalent to a planar graph, such as the dodecahedron shown in Fig.8.
Dodecahedron Planar graph
Fig.8
We have known a few criterions on planar graphs following.
Theorem 2.4(Euler,1758, [2]) Let G be a planar graph with p vertices, q edges and r faces.
Then, p− q + r = 2.
Theorem 2.5(Kuratowski,1930, [1]) A graph is planar if and only if it contains no subgraphs
homeomorphic with K5 or K(3, 3).
A 2-cell embedding of G on surface S is defined to be a continuous 1-1 mapping τ : G→ S
such that each component in S \ τ(G) homeomorphic to an open 2-disk { (x, y) | x2 + y2 < 1}.Certainly, the image τ(G) is contained in the 1-skeleton of a triangulation on the surface S.
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 7
For example, the embedding of K4 on the sphere and Klein bottle are shown in Fig.9 (a) and
(b), respectively.
bbbrrr r r r rr1 1
2
2
u1
u2
u3 u4
u1
u2u3
u46 6-�
Fig.9
There is an algebraic representation for characterizing the 2-cell embedding of graphs. For
v ∈ V (G), denote by NeG(v) = {e1, e2, · · · , eρ(v)} all the edges incident with the vertex v. A
permutation on e1, e2, · · · , eρ(v) is said to be a pure rotation and all pure rotations incident
with v is denoted by (v). Generally, a pure rotation system of the graph G is defined to be
ρ(G) = {(v)|v ∈ V (G)} which was observed and used by Dyck in 1888, Heffter in 1891 and
then formalized by Edmonds in 1960. For example,
ρ(K4) = {(u1u4, u1u3, u1u2), (u2u1, u2u3, u2u4), (u3u1, u3u4, u3u2), (u4u1, u4u2, u4u3)},ρ(K4) = {(u1u2, u1u3, u1u4), (u2u1, u2u3, u2u4), (u3u2, u3u4, u3u1), (u4u1, u4u2, u4u3)}
are respectively the pure rotation systems for embeddings of K4 on the sphere and Klein bottle
shown in Fig.9.
Theorem 2.6(Heffter 1891, Edmonds 1960, [4]) Every embedding of a graph G on an orientable
surface S induces a unique pure rotation system ρ(G). Conversely, Every pure rotation system
ρ(G) of a graph G induces a unique embedding of G on an orientable surface S.
Clearly, an embedding of graph G can be associated 0, 1 or 2-band respectively with
vertices, edges and face on its surface. A band decomposition is called locally orientable if each
0-band is assigned an orientation, and a 1-band is called orientation-preserving if the direction
induced on its ends by adjoining 0-bands are the same as those induced by one of the two
possible orientations of the 1-band. Otherwise, orientation-reversing. An edge e in a graph G
embedded on a surface S associated with a locally orientable band decomposition is said to be
type 0 if its corresponding 1-band is orientation-preserving and otherwise, type 1. A rotation
system ρL(v) of v ∈ V (G) to be a pair (J (v), λ), where J (v) is a pure rotation system and
λ : E(G) → Z2 is determined by λ(e) = 0 or λ(e) = 1 if e is type 0 or type 1 edge, respectively.
Theorem 2.7(Ringel 1950s, Stahl 1978,[4]) Every rotation system on a graph G defines a
unique locally orientable 2-cell embedding of G → S. Conversely, every 2-cell embedding of a
graph G→ S defines a rotation system for G.
8 Linfan MAO
A 2-cell embedding of a connected graph on surface is called map by W.T.Tutte. He
characterized embeddings by purely algebra in 1973 ([5], [26]). By his definition, an embedding
M = (Xα,β ,P) is defined to be a basic permutation P , i.e, for any x ∈ Xα,β, no integer k exists
such that Pkx = αx, acting on Xα,β , the disjoint union of quadricells Kx of x ∈ X (the base
set), where K = {1, α, β, αβ} is the Klein group, satisfying the following two conditions:
(1) αP = P−1α;
(2) the group ΨJ = 〈α, β,P〉 is transitive on Xα,β .
Furthermore, if the group ΨI = 〈αβ,P〉 is transitive on Xα,β , then M is non-orientable.
Otherwise, orientable.
For example, the embedding (Xα,β ,P) of graph K4 on torus shown in Fig.10.
--
6 6��� ZZZ rr rr?�}+�1
2 2
1
x
y
z
u
v
wsFig.10
can be algebraic represented by
Xα,β = {x, y, z, u, v, w, αx, αy, αz, αu, αv, αw, βx, βy,βz, βu, βv, βw, αβx, αβy, αβz, αβu, αβv, αβw},
P = (x, y, z)(αβx, u, w)(αβz, αβu, v)(αβy, αβv, αβw)
×(αx, αz, αy)(βx, αw, αu)(βz, αv, βu)(βy, βw, βv),
with vertices
v1 = {(x, y, z), (αx, αz, αy)}, v2 = {(αβx, u, w), (βx, αw, αu)},v3 = {(αβz, αβu, v), (βz, αv, βu)}, v4 = {(αβy, αβv, αβw), (βy, βw, βv)},
edges
{e, αe, βe, αβe}, e ∈ {x, y, z, u, v, w}
and faces
f1 = {(x, u, v, αβw, αβx, y, αβv, αβz), (βx, αz, αv, βy, αx, αw, βv, βu)},f2 = {(z, αβu,w, αβy), (βz, αy, βw, αu)}.
Two embeddings M1 = (X 1α,β ,P1) and M2 = (X 2
α,β ,P2) are said to be isomorphic if there
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 9
exists a bijection ξ
ξ : X 1α,β −→ X 2
α,β
such that for ∀x ∈ X 1α,β , ξα(x) = αξ(x), ξβ(x) = βξ(x), ξP1(x) = P2ξ(x). Particularly, if M1 =
M2 = M , an isomorphism between M1 and M2 is then called an automorphism of embedding
M . Clearly, all automorphisms of a embedding M form a group, called the automorphism group
of M , denoted by AutM .
There are two main problems on embedding of graphs on surfaces following.
Problem 2.1 Let G be a graph and S a surface. Whether or not G can be embedded on S?
This problem had been solved by Duke on orientable case in 1966, and Stahl on non-
orientable case in 1978. They obtained the result following.
Theorem 2.8(Duke 1966, Stahl 1978,[4]) Let G be a connected graph and let GR(G), CR(G)
be the respective genus range of G on orientable or non-orientable surfaces. Then, GR(G) and
CR(G) both are unbroken interval of integers.
Theorem 2.8 bring about to determine the minimum and maximum genus γ(G), γM (G) of
graph G on surfaces. Among them, the most simple case is to determine the maximum genus
γM (G) on non-orientable case, which was obtained by Edmonds in 1965. It is the Betti number
β(G) = |E(G)|− |V (G)|+1. The maximum genus γM (G) of G on orientable case is determined
by Xuong with the deficiency ξ(G), i.e., the minimum number of components in G \ T for all
spanning trees T in G in 1979. However, it is difficult for the minimum genus γ(G), only a few
results on typical graphs. For example, the genus of Kn and Km,n are listed following.
Theorem 2.9(Ringel and Youngs 1968, [4]) The minimum genus of a complete graph is given
by
γ(Kn) =
⌈(n− 3)(n− 4)
12
⌉, n ≥ 3.
Theorem 2.10(Ringel 1965, [4]) The minimum genus of a complete bipartite graph is given by
γ(K(m,n)) =
⌈(m− 2)(n− 2)
4
⌉, m, n ≥ 2.
Problem 2.2 Let G be a graph and S a surface. How many non-isomorphic embeddings of G
on S?
This problem is difficult, only be partially solved until today. However, the following simple
result enables one to enumerate rooted embeddings, where an embedding (Xα,β ,P) is rooted
on an element r ∈ Xα,β if r is marked beforehand.
Theorem 2.11([5],[6]) The auotmorphism group of a rooted embedding M , i.e., AutM r is
trivial.
Theorem 2.12([5],[6]) |AutM | | |Xα,β | = 4ε(M).
10 Linfan MAO
A root r in an embedding M is called an i-root if it is incident to a vertex of valency i. Two
i-roots r1, r2 are transitive if there exists an automorphism τ ∈ AutM such that τ(r1) = r2.
Define the enumerator v(D,x) and the root polynomials r(M,x), r(M(D), x) as follows:
v(D,x) =∑
i≥1
ivixi; r(M,x) =
∑
i≥1
r(M, i)xi,
where r(M, i) denotes the number of non-transitive i-roots in M .
Theorem 2.12 enables us to get the following results by applying the enumerator and root
polynomial of M .
Theorem 2.13(Mao and Liu, [21]) The number rO(Γ) of non-isomorphic rooted maps on ori-
entable surfaces underlying a simple graph Γ is
rO(Γ) =
2ε(Γ)∏
v∈V (Γ)
(ρ(v) − 1)!
|AutΓ| ,
where ε(Γ),ρ(v) denote the size of Γ and the valency of the vertex v, respectively.
Theorem 2.14(Mao and Liu, [22]) The number rN (Γ) of rooted maps on non-orientable surfaces
underlying a graph Γ is
rN (Γ) =
(2β(Γ)+1 − 2)ε(Γ)∏
v∈V (Γ)
(ρ(v) − 1)!
∣∣∣Aut 12Γ∣∣∣
.
For a few well-known graphs, Theorems 2.13 and 2.14 enables us to get Table 2.
G rO(G) rN (G)
Pn n− 1 0
Cn 1 1
Kn (n− 2)!n−1 (2(n−1)(n−2)
2 − 1)(n− 2)!n−1
Km,n(m 6= n) 2(m− 1)!n−1(n− 1)!m−1 (2mn−m−n+2 − 2)(m− 1)!n−1(n− 1)!m−1
Kn,n (n− 1)!2n−2 (2n2−2n+2 − 1)(n− 1)!2n−2
Bn(2n)!2nn! (2n+1 − 1) (2n)!
2nn!
Dpn (n− 1)! (2n − 1)(n− 1)!
Dpk,ln (k 6= l) (n+k+l)(n+2k−1)!(n+2l−1)!
2k+l−1n!k!l!(2n+k+l−1)(n+k+l)(n+2k−1)!(n+2l−1)!
2k+l−1n!k!l!
Dpk,kn
(n+2k)(n+2k−1)!2
22kn!k!2(2n+2k−1)(n+2k)(n+2k−1)!2
22kn!k!2
Table 2
Apply the Burnside Lemma in permutation groups, we got the numbers of unrooted maps
of complete graph Kn on orientable or non-orientable surfaces by calculating the stabilizer of
each automorphism of complete maps.
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 11
Theorem 2.15(Mao, Liu and Tian, [23]) The number nO((Kn) of complete maps of order
n ≥ 5 on orientable surfaces is
nO(Kn) =1
2(∑
k|n+
∑
k|n,k≡0(mod2)
)(n− 2)!
nk
knk (n
k)!
+∑
k|(n−1),k 6=1
φ(k)(n − 2)!n−1
k
n− 1.
and n(K4) = 3.
Theorem 2.16(Mao, Liu and Tian, [23]) The number nN (Kn) of complete maps of order
n, n ≥ 5 on non-orientable surfaces is
nN (Kn) =1
2(∑
k|n+
∑
k|n,k≡0(mod2)
)(2α(n,k) − 1)(n− 2)!
nk
knk (n
k)!
+∑
k|(n−1),k 6=1
φ(k)(2β(n,k) − 1)(n− 2)!n−1
k
n− 1,
and nN (K4) = 8, where,
α(n, k) =
n(n − 3)
2k, if k ≡ 1(mod2);
n(n − 2)
2k, if k ≡ 0(mod2),
β(n, k) =
(n − 1)(n − 2)
2k, if k ≡ 1(mod2);
(n − 1)(n − 3)
2k, if k ≡ 0(mod2).
§3. Complex Networks with Reality
A network is a directed graph G associated with a non-negative integer-valued function c on
edges and conserved at each vertex, which are abstracting of practical networks, for instance,
the electricity, communication and transportation networks such as those shown in Fig.11 for
the high-speed rail network in China planed a few years ago.
Fig.11
Clearly, a network is nothing else but a labeled graph GL with L : E(G) → Z+. generally,
a labeled graph on a graph G = (V,E; I) is a mapping θL : V ∪E → L for a label set L, denoted
by GL. If θL : E → ∅ or θL : V → ∅, then GL is called a vertex labeled graph or an edge labeled
12 Linfan MAO
graph, denoted by GV or GE , respectively. Otherwise, it is called a vertex-edge labeled graph.
Similarly, two networks GL11 , GL2
2 are equivalent, if there is an isomorphism φ : G1 → G2 such
that φ(L1(x)) = L2(φ(x)) for x ∈ V (G1)⋃E(G2).
It should be noted that labeled graphs are more useful in understanding matters in the
universe. For example, there is a famous story, i.e., the blind men with an elephant. In this
story, 6 blind men were be asked to determine what is an elephant looks like. The man touched
the elephant’s leg, tail, trunk, ear, belly or tusk respectively claims it’s like a pillar, a rope,
a tree branch, a hand fan, a wall or a solid pipe. Each of them insisted on his own and not
accepted others.
Fig.12
They then entered into an endless argument. All of you are right! A wise man explained to
them: why are you telling it differently is because each one of you touched the different part
of the elephant. What is the meaning of the wise man? He claimed nothing else but the looks
like of an elephant, i.e.,
An elephant = {4 pillars}⋃
{1 rope}⋃
{1 tree branch}⋃ {2 hand fans}
⋃{1 wall}
⋃{1 solid pipe}.
Usually, a thing T is identified with known characters on it at one time, and this process is
advanced gradually by ours. For example, let µ1, µ2, · · · , µn be the known and νi, i ≥ 1 the
unknown characters at time t. Then, the thing T is understood by
T =
(n⋃
i=1
{µi})⋃⋃
k≥1
{νk}
in logic and with an approximation T ◦ =n⋃
i=1
{µi} at time t, which are both Smarandache
multispace ([7],[26]).
What is the implications of this story for understanding matters in the universe? It lies
in the situation that humans knowing matters in the universe is analogous to these blind men.
However, if the wise man were L.Euler, a mathematician he would tell these blind men that an
elephant looks like nothing else but a tree labeled by sets as shown in Fig.13, where, {a} =tusk,
{b1, b2} =ears, {c} =head, {d} =belly, {e1, e2, e3, e4} =legs and {f} =tail with their intersection
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 13
sets labeled on edges.l lll l ll ll la c d f
b1
b2
e1
e2
e3
e4
Fig.13
For the case of Euclidean space with dimension≥ 3, the intuition tells us that to embed
a graph in a of dimension is not difficult by the result following. However, it is obvious but a
universal skeleton inherited in all matters.
Theorem 3.1 A simple graph G can be rectilinear embedded, i.e., all edge are segment of
straight line in a Euclidean space Rn with n ≥ 3.
In fact, we can choose n district points in curve (t, t2, t3) of Euclidean space R3 on n different
values of t. Then, it can be easily show that all these straight lines are never intersecting.
Whence, it is a trivial problem on embedding graphs of R3. However, all matters are in 3-
dimensional Euclidean space in the eyes of humans, i.e., the reality of a matter in the universe
should be understanding on its 1-dimensional skeleton in the space.
Then, what is the reality of a matter? The word reality of a matter T is its state as it actually
exist, including everything that is and has been, no matter it is observable or comprehensible
by humans. How can we hold on the reality of matters? Usually, a matter T is multilateral,
i.e., Smarandache multispace or complex one and so, hold on its reality is difficult for humans
in logic, such as the meaning in the story of the blind men with an elephant.
For hold on the reality of matters, a general notion is
MatterDecompose−→ Microcosmic Particles
Abstract−→ Complex Network.
For example, the physics determine the nature of matters by subdividing a matter to an irre-
ducibly smallest detectable particle ([28]), i.e., elementary particles, which is essentially transfer
the matter to a complex network such as those meson’s and baryon’s composition by quarks.
Similarly, the basic unit of life or the basic unit of heredity are cells and genes in biology
which also enables us to get the life networks of cell or genes. This notion can be found in all
modern science with an conclusion that a matter = a complex network. Its essence of this notion
is to determine the nature of irreducibly smallest detectable units and then, holds on reality
of the matter. However, a matter can be always divided into submatters, then sub-submatters
and so on. A natural question on this notion is whether it has a terminal point or not. On the
other hand, it is a very large complex network in general. For example, the complex network of
a human body consists of 5×1014−−6×1014 cells. Are we really need such a large and complex
network for the reality of matters? Certainly not! How can we hold on the reality of matters
14 Linfan MAO
by such a complex network? And do we have mathematical theory on complex network? The
answer is not certain because although we have established a theory on complex network but
it is only a local theory by combination of the graphs and statistics with the help of computer
([3]), can not be used for the reality of matters.
However, we find a beacon light inspired by the traditional Chinese medicine. There are
12 meridians which completely reflects the physical condition of human body in traditional
Chinese medicine: LU, LI, ST, SP, HT, SI, BL, KI, PC, SJ, GB, LR. For example, the LI and
GB meridians are shown in Fig.14.
Fig.14
All of these 12 meridians can be classified into 3 classes following:
Class 1. Paths, including LU, LI, SP, HT, SI, KI, PC and LR meridians;
Class 2. Trees, including GB, ST and SJ meridians;
Class 3. Gluing Product of circuit with paths Cn
⊙Pm1
⊙Pm2 , including BL meridian.
According to the Standard China National Standard (GB 12346-90), the inherited graph
of the 12 meridians on a human body is shown in Fig.15.q qq qq
qq q qqq
qqqq qq
qST1
ST8
ST5
ST45
BL10
GB15
BL40
GB44BL67
q q qqqqq qqq q qKI27 LU1
SP6
LR1 SP1KI1
qqHT1
HT9
LR14PC1
SP21
q q qqq qPC9
LU11
BL1
qq q qq
q qqSJ23
SJ20 SJ16
SJ1
qq q qq
GB1
SI18
SI1
SI19qq qqq q qq qqq q
q qqqqq
�� qq qq qqqqq q qq q
qqqq qqq qq qqq qqqq
qq qqq
qFig.15 12 Meridian graph on a human body
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 15
By the traditional Chinese medicine ([28]), if there exists an imbalanced acupoint on one of
the 12 meridians, this person must has illness and in turn, there must be imbalance acupoints
on the 12 meridians for a patient. Thus, finding out which acupoint on which meridian is in
imbalance with Yin more than Yang or Yang more than Yin is the main duty of a Chinese
doctor. Then, the doctor regulates the meridian by acupuncture or drugs so that the balance
on the imbalance acupoints recovers again, and then the patient recovers.
Then, what is the significance of the treatment theory in traditional Chinese medicine
to science? It implies we are not need a large complex network for holding on the body of
human. Whether or not classically mathematical elements enough for understanding complex
networks, i.e., matters in the universe? The answer is negative because all of them are local.
Then, could we establish a mathematics over elements underlying combinatorial structures? The
answer is affirmative, i.e., mathematical combinatorics discussed in this paper. Certainly, we
can introduce 2 kinds of elements respectively on living of non-living matters.
Element 1(Non-Living Body). A continuity flow−→GL is an oriented embedded graph−→
G in a topological space S associated with a mapping L : v → L(v), (v, u) → L(v, u), 2
end-operators A+vu : L(v, u) → LA+
vu(v, u) and A+uv : L(u, v) → LA+
uv (u, v) on a Banach space B
over a field F such as those shown in Fig.16,-���� ����L(v, u)A+vu A+
uvL(v) L(u)
v uFig.16
with L(v, u) = −L(u, v), A+vu(−L(v, u)) = −LA+
vu(v, u) for ∀(v, u) ∈ E(−→G)
and holding with
continuity equation ∑
u∈NG(v)
LA+vu (v, u) = L(v) for ∀v ∈ V
(−→G).
Element 2(Living Body). A harmonic flow−→GL is an oriented embedded graph
−→G in a
topological space S associated with a mapping L : v → L(v) − iL(v) for v ∈ E(−→G)
and L :
(v, u) → L(v, u)− iL(v, u), 2 end-operators A+vu : L(v, u)− iL(v, u) → LA+
vu(v, u)− iLA+vu(v, u)
and A+uv : L(v, u) − iL(v, u) → LA+
uv(v, u) − iLA+uv(v, u) on a Banach space B over a field F
such as those shown in Fig.17, -���� ����L(v, u) − iL(v, u)A+vu A+
uvL(v) L(u)
v uFig.17
where i2 = −1, L(v, u) = −L(u, v) for ∀(v, u) ∈ E(−→G)
and holding with continuity equation
∑
u∈NG(v)
(LA+
vu (v, u) − iLA+vu (v, u)
)= L(v) − iL(v)
for ∀v ∈ V(−→G).
16 Linfan MAO
Let G be a closed family of graphs−→G under the union operation and let B be a linear
space (B; +, ·), or furthermore, a commutative ring, a Banach or Hilbert space (B; +, ·) over a
field F . Denoted by (GB; +, ·) and(G
±B
; +, ·)
the respectively elements 1 and 2 form by graphs
G ∈ G . Then, elements 1 and 2 can be viewed as vectors underlying an embedded graph G in
space, which enable us to establish mathematics on complex networks and get results following.
Theorem 3.2([9-10,14-18]) If G is a closed family of graphs−→G under the union operation and
B a linear space (B; +, ·), then, (GB; +, ·) and(G
±B
; +, ·)
with linear operators A+vu, A+
uv for
∀(v, u) ∈ E
( ⋃G∈G
−→G
)under operations + and · form respectively a linear space, and further-
more, a commutative ring if B is a commutative ring (B; +, ·) over a field F .
Theorem 3.3([9-10,14-18]) If G is a closed family of graphs under the union operation and
B a Banach space (B; +, ·), then, (GB; +, ·) and(G
±B
; +, ·)
with linear operators A+vu, A+
uv for
∀(v, u) ∈ E
( ⋃G∈G
−→G
)under operations + and · form respectively a Banach or Hilbert space
respect to that B is a Banach or Hilbert space.
A few well-known results such as those of Banach theorem, closed graph theorem and
Hahn-Banach theorem are also generalized on elements 1 and 2. For example, we obtained
results following.
Theorem 3.4(Taylor, [15]) Let−→GL ∈
⟨−→G i, 1 ≤ i ≤ n
⟩R×Rn
and there exist kth order derivative
of L to t on a domain D ⊂ R, where k ≥ 1. If A+vu, A+
uv are linear for ∀(v, u) ∈ E(−→G), then
−→GL =
−→GL(t0) +
t− t01!
−→GL′(t0) + · · · + (t− t0)
k
k!
−→GL(k)(t0) + o
((t− t0)
−k −→G),
for ∀t0 ∈ D , where o((t− t0)
−k −→G)
denotes such an infinitesimal term L of L that
limt→t0
L(v, u)
(t− t0)k
= 0 for ∀(v, u) ∈ E(−→G).
Particularly, if L(v, u) = f(t)cvu, where cvu is a constant, denoted by f(t)−→GLC with LC :
(v, u) → cvu for ∀(v, u) ∈ E(−→G)
and
f(t) = f(t0) +(t − t0)
1!f′(t0) +
(t − t0)2
2!f′′(t0) + · · · + (t − t0)
k
k!f
(k)(t0) + o((t − t0)
k)
,
then
f(t)−→GLC = f(t) · −→GLC .
Theorem 3.5(Hahn-Banach, [19]) Let H±
Bbe an element 2 subspace of G
±B
and let F : H±
B→
C be a linear continuous functional on H±
B. Then, there is a linear continuous functional
F : G±B
→ C hold with
Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory 17
(1) F(−→GL2
)= F
(−→GL2
)if−→GL2 ∈ H
±B
;
(2)∥∥∥F∥∥∥ = ‖F‖.
For applications of elements 1 and 2 to physics and other sciences such as those of ele-
mentary particles, gravitations, ecological system, · · · etc., the reader is refereed to references
[11]-[13] and [18]-[20] for details.
References
[1] J.A.Bondy and U.S.R.Murty, Graph Theory with Applications, The Macmillan Press Ltd,
1976.
[2] G.Chartrand and L.Lesniak, Graphs & Digraphs, Wadsworth, Inc., California, 1986.
[3] G.R.Chen, X.F.Wang and X.Li, Introduction to Complex Networks – Models, Structures
and Dynamics (2 Edition), Higher Education Press, Beijing, 2015.
[4] J.L.Gross and T.W.Tucker, Topological Graph Theory, John Wiley & Sons,1987.
[5] Y.P.Liu, Enumerative Theory of Maps, Kluwer Academic Publishers , Dordrecht/ Boston/
London,1999.
[6] Linfan Mao, Automorphism Groups of Maps, Surfaces and Smarandache Geometries (Sec-
ond edition), The Education Publisher Inc., USA, 2011.
[7] Linfan Mao, Smarandache Multi-Space Theory, The Education Publisher Inc., USA, 2011.
[8] Linfan Mao, Combinatorial Geometry with Applications to Field Theory, The Education
Publisher Inc., USA, 2011.
[9] Linfan Mao, Mathematics on non-mathematics - A combinatorial contribution, Interna-
tional J.Math. Combin., Vol.3(2014), 1-34.
[10] Linfan Mao, Extended Banach−→G -flow spaces on differential equations with applications,
Electronic J.Mathematical Analysis and Applications, Vol.3, No.2 (2015), 59-91.
[11] Linfan Mao, A new understanding of particles by−→G -flow interpretation of differential
equation, Progress in Physics, Vol.11(2015), 193-201.
[12] Linfan Mao, A review on natural reality with physical equation, Progress in Physics,
Vol.11(2015), 276-282.
[13] Linfan Mao, Mathematics with natural reality – Action Flows, Bull.Cal.Math.Soc., Vol.107,
6(2015), 443-474.
[14] Linfan Mao, Mathematical combinatorics with natural reality, International J.Math. Com-
bin., Vol.2(2017), 11-33.
[15] Linfan Mao, Hilbert flow spaces with operators over topological graphs, International
J.Math. Combin., Vol.4(2017), 19-45.
[16] Linfan Mao, Complex system with flows and synchronization, Bull.Cal.Math.Soc., Vol.109,
6(2017), 461-484.
[17] Linfan Mao, Mathematical 4th crisis: to reality, International J.Math. Combin., Vol.3(2018),
147-158.
[18] Linfan Mao, Harmonic flow’s dynamics on animals in microscopic level with balance re-
covery, International J.Math. Combin., Vol.1(2019), 1-44.
18 Linfan MAO
[19] Linfan Mao, Science’s dilemma – A review on science with applications, Progress in Physics,
Vol.15(2019), 78–85.
[20] Linfan Mao, A new understanding on the asymmetry of matter-antimatter, Progress in
Physics, Vol.15, 3(2019), 78-85.
[21] Linfan Mao and Yanpei Liu, On the roots on orientable embeddings of graph, Acta Math.
Scientia, 23A(3), 287-293(2003).
[22] Linfan Mao and Yanpei Liu, Group action approach for enumerating maps on surfaces, J.
Applied Math. & Computing, Vol.13, No.1-2,201-215.
[23] Linfan Mao,Yanpei Liu and Feng Tian, The number of complete maps on surfaces, arXiv:
math.GM/0607790. Also in L.Mao ed., Selected Papers on Mathematical Combinatorics(I),
World Academic Union, 2006, 94-118.
[24] W.S.Massey, Algebraic topology: An introduction, Springer-Verlag, New York, etc., 1977.
[25] Quang Ho-Kim and Pham Xuan Yem, Elementary Particles and Their Interactions, Springer-
Verlag Berlin Heidelberg, 1998.
[26] F.Smarandache, Paradoxist Geometry, State Archives from Valcea, Rm. Valcea, Romania,
1969, and in Paradoxist Mathematics, Collected Papers (Vol. II), Kishinev University
Press, Kishinev, 5-28, 1997.
[27] W.T.Tutte, What is a map? in New Directions in the Theory of Graphs, ed. by F.Harary,
Academic Press, 1973, 309-325.
[28] Zhichong Zhang, Comments on the Inner Canon of Emperor(Qing Dynasty, in Chinese),
Northern Literature and Art Publishing House, 2007.
International J.Math. Combin. Vol.4(2019), 19-28
Some Subclasses of
Meromorphic with p-Valent q-Spirallike Functions
Read.S.A.Qahtan, Hamid Shamsan and S.Latha
(Department of Mathematics, Yuvaraja’s College, University of Mysore, Mysore 570005, India)
E-mail: [email protected], [email protected], [email protected]
Abstract: In this paper, we introduce and investigate two new subclasses of the function
class ΣMS(p, λ, β, q) and ΣMC(p, λ, β, q) of q-spirallike meromorphic functions defined in
the punctured open unit disc.
Key Words: Univalent functions, meromorphic functions, meromorphic q-spirallike func-
tions.
AMS(2010): 30C45.
§1. Introduction
Let Σp be the class of functions f of the form
f(z) =1
zp+
∞∑
n=1−p
an zn, (p ∈ N = 1, 2, 3, · · · ), (1)
which are analytic in the open disc E∗ = {z : z ∈ C and 0 < |z| < 1}. Let S be the subclass of
functions in Σp which are univalent in E. Let P be the class of functions p given by
p(z) = 1 +
∞∑
n=1
pn zn (z ∈ E), (2)
which are analytic in the open disc E and satisfy the condition:
ℜ{p(z)} > 0 (z ∈ E). (3)
If f ∈ Σp and satisfies
−ℜ{zf ′(z)
f(z)
}> β (z ∈ E, 0 ≤ β < p), (4)
then we say that f is meromorphic p-valent starlike of order β (0 ≤ β < p) and we denote this
class by ΣMS(p, β).
1Received July 11, 2019, Accepted November 22, 2019.
20 Read.S.A.Qahtan, Hamid Shamsan and S.Latha
If f ∈ Σp and satisfies
−ℜ{
1 +zf ′′(z)
f ′(z)
}> β (z ∈ E, 0 ≤ β < p), (5)
then we say that f is meromorphic p-valent convex of order β and we denote this class by
f ∈ ΣMC(p, β).
A function f ∈ Σp is said to be λ-spirallike of order β in the unit disk E if
−ℜ{eiλ zf
′(z)
f(z)
}> β, (z ∈ E, 0 ≤ β < p, |λ| < π
2).
In [8] Jackson introduced and studied the concept of the q-derivative operator ∂q as
follows
∂qf(z) =f(z) − f(qz)
z (1 − q), (z 6= 0, 0 < q < 1, ∂qf(0) = f ′(0)). (6)
Equivalently (6) may be written as
∂qf(z) = 1 +
∞∑
n=2
[n]q an zn−1, z 6= 0, (7)
where [n]q = 1− qn
1− q. Note that as q → 1−, [n]q → n.
Definition 1.1 A function f ∈ Σp is said to be meromorphic p-valent λ-q-spirallike functions
of order β if it satisfies the following:
−ℜ{eiλ z∂qf(z)
f(z)
}> β (z ∈ E, |λ| < π
2, 0 ≤ β > p cosλ, 0 < q ≤ 1), (8)
we denote this class by ΣMS(p, λ, β, q).
Definition 1.2 A function f ∈ Σp is said to be meromorphic p-valent convex λ-q-spirallike
functions of order β if it satisfies the following
−ℜ{eiλ ∂q(z∂qf(z))
∂qf(z)
}> β (z ∈ E, 0 ≤ β < 1), (9)
we denote this class by ΣMC(p, λ, β, q).
Remark 1.1 f ∈ MS(p, λ, β, q) iff
−eiλ z∂qf(z)
f(z)≺ peiλ − (2β − pe−iλ)z
1 − z, (10)
and f ∈ MC(p, λ, β, q) iff
−eiλ
(∂q(z∂qf(z))
∂qf(z)
)≺ peiλ − (2β − pe−iλ)z
1 − z. (11)
Some Subclasses of Meromorphic with p-Valent q-Spirallike Functions 21
§2. Main Results
Theorem 2.1 If the sequence {Ap+m}∞0 defined by
Ap = 2(β−p cos λ)p+[p]q
, m = 0,
Ap+m = 2(β−p cos λ)p+[p+m]q
(1 +
∑m−1k=0 |ap+k|
), m ∈ N ,
(12)
and p ∈ N. Then
Ap+m =2(β − p cosλ)
2β + [m+ p]q + p− 2p cosλ
m∏
k=0
2β + [k + p]q + p− 2p cosλ
p+ [p+ k]q, (13)
where m ∈ N0 = N \ {0}.
Proof By virtue of (12), we have
p+ [p+m+ 1]qAp+m+1 = 2(β − p cosλ)
(1 +
m∑
k=0
Ap+k
), (14)
and
p+ [p+m]qAp+m = 2(β − p cosλ)
(1 +
m−1∑
k=0
Ap+k
). (15)
From (14) and (15), we have
Ap+m+1
Ap+m
=2β + [m+ p]q + p− 2p cosλ
p+ [p+m+ 1]qm ∈ N0. (16)
Ap+m =Ap+m
Ap+m−1.Ap+m−1
Ap+m−2· · · Ap+1
Ap
.Ap
=2β + [m+ p− 1]q + p− 2p cosλ
p+ [p+m]q· · · 2β + [p]q + p− 2p cosλ
p+ [p+ 1]q.2β − 2p cosλ
p+ [p]q
=2(β − p cosλ)
2β + [m+ p]q + p− 2p cosλ
m∏
k=0
2β + [k + p]q + p− 2p cosλ
p+ [p+ k]q(m ∈ N).
(17)
The proof of Theorem 2.1 is completed. 2As q → 1−, we get the following result proved by Shi.et al. [13].
Corollary 2.1 If {Ap+m}∞0 defined by
Ap = β−p cos λ
p, m = 0,
Ap+m = 2(β−p cos λ)2p+m
(1 +
∑m−1k=0 |ap+k|
), m ∈ N ,
(18)
22 Read.S.A.Qahtan, Hamid Shamsan and S.Latha
and p ∈ N. Then
Ap+m ≤ 2(β − p cosλ)
2β +m+ 2p− 2p cosλ
m∏
k=0
2β + k + 2p− 2p cosλ
2p+ k, (19)
where, m ∈ N0 = N \ {0}.
Theorem 2.2 Let f(z) = 1zp +
∑∞m=0 ap+mz
p+m ∈ MS(p, λ, β, q). Then
|ap+m| ≤ 2(β − p cosλ)
2β + [m+ p]q + p− 2p cosλ
m∏
k=0
2β + [k + p]q + p− 2p cosλ
p+ [p+ k]q(m ∈ N0). (20)
Proof Let
L(z) =β + eiλ z∂qf(z)
f(z) + ip sinλ
β − p cosλ(z ∈ E, f ∈ MS(p, λ, β, q)). (21)
We know that L ∈ P . It follows that
eiλz∂qf(z) = (β − p cosλ)f(z)L(z) − (β − ip sinλ)f(z). (22)
Let
L(z) = 1 + l1z + l2z2 + · · · . (23)
Then
eiβ
(−[p]qzp
+ [p]qapzp + [p+ 1]qap+1z
p+1 + · · · + [p+m]qap+mzp+m + · · ·
)
= (β − p cosλ)
(1
zp+ apz
p + ap+1zp+1 + · · ·
)× (1 + l1z + l2z
2 + · · · )
− (β − ip sinλ)
(1
zp+ apz
p + ap+1zp+1 + · · · + ap+mz
p+m + · · ·).
(24)
We have from sides (24)
eiλ[p+m]qap+m = (β − p cosλ) (l2p+m + aplm + ap+mlm−1
+ · · · +ap+m) − (β + ip sinλ)ap+m. (25)
Moreover, we know that
|ln| ≤ 2 (n ∈ N). (26)
From (25) and(26) we have
|ap| ≤2(β − p cosλ)
p+ [p]q(27)
Some Subclasses of Meromorphic with p-Valent q-Spirallike Functions 23
and
|ap+m| ≤ 2(β − p cosλ)
p+ [p+m]q
(1 +
m−1∑
k=0
|ap+k|)
(28)
with supposing p ∈ N . We define {Ap+m}∞m=0 by
Ap = 2(β−p cos λ)
p+[p]q, m = 0;
Ap+m = 2(β−p cos λ)p+[p+m]q
(1 +
∑m−1k=0 |ap+k|
), m ≥ 1.
(29)
Now by the mathematical induction principle we will prove that
|ap+m| ≤ Ap+m(m ∈ N0). (30)
We can easily verify that
|ap| ≤ Ap =2(β − p cosλ)
p+ [p]q. (31)
Thus, assuming that
|ap+j | ≤ Ap+j(j = 0, 1, · · · ,m.m ∈ N0), (32)
from (28)) and (32) we have
|ap+m+1| ≤2(β − p cosλ)
p+ [p+m+ 1]q
(1 +
m∑
k=0
|ap+k|)
≤ 2(β − p cosλ)
p+ [p+m+ 1]q
(1 +
m∑
k=0
|Ap+k|)
=Ap+m+1 (m ∈ N0).
(33)
Therefore, by the principle of mathematical induction, we have
|ap+m| ≤ Ap+m (m ≤ N0). (34)
By means of Theorem 2.1 and (29), we know that
Ap+m =2(β − p cosλ)
2β + [m+ p]q + p− 2p cosλ
m∏
k=0
2β + [k + p]q + p− 2p cosλ
p+ [p+ k]q(m ∈ n0). (35)
So, from (34) and (35) we get proof of the Theorem 2.2. 2As q → 1− we get the following result proved by Shi.et al. [13].
Corollary 2.2 Let f(z) = 1zp +
∑∞m=0 ap+mz
p+m ∈ MS(p, λ, β). Then
Ap+m =2(β − p cosλ)
2β +m+ 2p− 2p cosλ
m∏
k=0
2β + k + 2p− 2p cosλ
2p+ k(m ∈ n0). (36)
24 Read.S.A.Qahtan, Hamid Shamsan and S.Latha
From Theorem 2.2 we get the following result.
Corollary 2.3 Let f(z) = 1zp +
∑∞m=0 ap+mz
p+m ∈ MC(p, λ, β, q). Then
Ap+m =2p(β − p cosλ)
[p+m](2β + [m+ p]q + p− 2p cosλ)
m∏
k=0
2β + [k + p]q + p− 2p cosλ
p+ [p+ k]q(m ∈ n0). (37)
Theorem 2.3 Let f(z) = 1zp +
∑∞m=0 ap+mz
p+m ∈ MS(p, λ, β, q). Then
p cosλ− 2(β − p cosλ)r
1 − r≤ ℜ
(−eiϑ z∂qf(z)
f(z)
)≤ p cosλ+ 2(β − p cosλ)r
1 + r(38)
for |z| = r < 1.
Proof Suppose the function φ defined by
φ(z) =peiλ − (2β − pe−iλz)
1 − z(z ∈ E). (39)
Let z = reiλ (0 < r < 1). We have
ℜ{φ(z)} = p cosλ− 2(β − p cosλ)r(cos ϑ− r)
1 + r2 − 2r cosϑ. (40)
Let
ϕ(τ) = p cosλ− 2(β − p cosλ)r(τ − r)
1 + r2 − 2τr(τ = cosϑ). (41)
Then
∂qϕ(τ) =−2r(β − p cosλ)[(1 + r2 − 2τr) − r[2r]q(τ − r)]
(1 + r2 − 2qrτ)(1 + r2 − 2rτ). (42)
This means that
p cosλ− 2(β − p cosλ)r
1 − r≤ ℜ(φ(z)) ≤ p cosλ+
2(β − p cosλ)r
1 + r, (43)
which is equivalent to
p cosλ− 2(β − p cosλ)r
1 − r≤ ℜ{φ(z)} ≤ p cosλ+ 2(β − p cosλ)r
1 + r. (44)
We know that
−eiλ z∂qf(z)
f(z)≺ φ(z)
and φ(z) is univalent in E, this is prove the inequality (38). 2As q → 1− we get the following result proved by Shi.et al. [13].
Some Subclasses of Meromorphic with p-Valent q-Spirallike Functions 25
Corollary 2.4 Let f(z) = 1zp +
∑∞m=0 ap+mz
p+m ∈ MS(p, λ, β). Then
p cosλ− 2(β − p cosλ)r
1 − r≤ ℜ
(−eiλ zf
′(z)
f(z)
)≤ p cosλ+ 2(β − p cosλ)r
1 + r, (45)
for |z| = r < 1.
Corollary 2.5 Let f(z) = 1zp +
∑∞m=0 ap+mz
p+m ∈ MC(p, λ, β, q). Then
p cosλ− 2(β − p cosλ)r
1 − r≤ ℜ
(−eiλ ∂q(z∂qf(z))
∂qf(z)
)≤ p cosλ+ 2(β − p cosλ)r
1 + r(46)
for |z| = r < 1.
Theorem 2.4 If f ∈ Σp satisfies
q∞∑
n=1−p
(|[n]qe
iλ + γ| + |[n]qeiλ + 2β − γ|
)|an| ≤ |[p]qeiλ − 2qβ + qγ| − |[p]qeiλ − qγ| (47)
for some real λ, β and γ (0 ≤ γ ≤ p cosλ), then f ∈ MS(p, λ, β, q)
Proof To prove f ∈ MS(p, λ, β, q), it suffices to show that
∣∣∣∣∣∣
eiλ z∂qf(z)f(z) + γ
eiλ z∂qf(z)f(z) + (2β − γ)
∣∣∣∣∣∣< 1 (z ∈ E, 0 ≤ γ ≤ p cosλ). (48)
From (47), we know that
∣∣[p]qeiλ − 2qβ + qγ∣∣+ q
∞∑
n=1−p
∣∣[n]qeiλ + 2β − γ
∣∣ |an| ≥∣∣[p]qeiλ − qγ
∣∣
+q
∞∑
n=1−p
|[n]qeiλ + γ||an| > 0. (49)
Now, by the maximum modulus principle, we deduce from (1) and (49) that
∣∣∣∣∣∣
eiλ z∂qf(z)f(z) + γ
eiλ z∂qf(z)f(z) + (2β − γ)
∣∣∣∣∣∣=
∣∣∣∣∣∣
eiλ(
−[p]qqzp +
∑∞n=1−p[n]qanz
n)
+ γzp + γ
∑∞n=1−p anz
n
eiλ
(−[p]qqzp +
∑∞n=1−p[n]qanzn
)+ (2β − γ)
(1zp +
∑∞n=1−p anzn
)
∣∣∣∣∣∣
=
∣∣∣∣∣(−[p]qe
iλ + qγ) + q∑∞
n=1−p([n]qeiλ + γ)anz
n
(−[p]qeiλ + 2qβ − qγ) + q∑∞
n=1−p([n]qeiλ + 2β − γ)anzn
∣∣∣∣∣
<|[p]qeiλ − qγ| + q
∑∞n=1−p |[n]qe
iλ + γ||an||[p]qeiλ − 2qβ + qγ| − q
∑∞n=1−p |[n]qeiλ + 2β − γ||an|
≤1.
(50)
This completes the proof. 2
26 Read.S.A.Qahtan, Hamid Shamsan and S.Latha
As q → 1− we get the following result proved by Shi.et al.[13].
Corollary 2.6 If f ∈ Σp satisfies the
∞∑
n=1−p
(|neiλ + γ| + |neiλ + 2β − γ|
)|an| ≤ |peiλ − 2β + γ| − |peiλ − γ| (51)
for some real λ, β and γ (0 ≤ γ ≤ p cosλ), then f ∈ MC(p, λ, β).
Corollary 2.7 If f ∈∑@p satisfies the
q
∞∑
n=1−p
|[n]q|(|[n]qe
iλ + γ| + |[n]qeiλ + 2β − γ|
)|an| ≤ [p]q
(|[p]qeiλ − 2qβ + qγ|
− |[p]qeiλ − qγ|)
(52)
for some real λ, β and γ (0 ≤ γ ≤ p cosλ), then f ∈ MC(p, λ, β, q).
Lemma 2.1([7]) If |φ| attains its maximum value on the circle |z| = r < 1 at z0 and φ is a
nonconstant regular function in E then
z0φ′(z0) = kφ(z0), k ≥ 1, k ∈ R.
Theorem 2.5 If f ∈ Σp satisfies
∣∣∣∣∣f(z)
f(qz)+zf(z)∂2
qf(z)
f(qz)∂qf(z)− z∂qf(z)
f(qz)
∣∣∣∣∣ <β − p
2β(53)
for some real β > p, than f ∈ MS(p, 0, β, q).
Proof Define the function ϕ by
ϕ(z) =
z∂qf(z)f(z) + p
z∂qf(z)f(z) + (2β − p)
(z ∈ E). (54)
Note that ϕ is analytic in E and ϕ(0) = 0. From (54), we have
z∂qf(z)
f(z)=
−p+ (2β − p)ϕ(z)
1 − ϕ(z). (55)
Taking q-differentiating of (55) logarithmically, we get
f(z)
f(qz)+zf(z)∂2
qf(z)
f(qz)∂qf(z)− z∂qf(z)
f(qz)=
z(1 − ϕ(z))(2β − p)∂qϕ(z)
(−p+ (2β − p)ϕ(z))(1 − ϕ(qz))+
z∂qϕ(z)
(1 − ϕ(qz)). (56)
Some Subclasses of Meromorphic with p-Valent q-Spirallike Functions 27
From (53) and (56), we get that
∣∣∣∣∣f(z)
f(qz)+zf(z)∂2
qf(z)
f(qz)∂qf(z)− z∂qf(z)
f(qz)
∣∣∣∣∣ =
∣∣∣∣2(β − p)∂qϕ(z)
[−p+ (2β − p)ϕ(z)](1 − ϕ(qz))
∣∣∣∣ <β − p
2β. (57)
Consider z0 ∈ E such that
max|z|≤|z0|
|ϕ(z)| = |ϕ(z0)| = 1.
By Lemma 2.1, let ϕ(z0) = eiϑ and z0∂qϕ(z0) = Leiϑ (L ≥ 1). For such a point z0, we have
that ∣∣∣∣∣f(z0)
f(qz0)+z0f(z0)∂
2qf(z0)
f(qz0)∂qf(z0)− z0∂qf(z0)
f(qz0)
∣∣∣∣∣
=
∣∣∣∣2(β − p)Leiϑ
[−p+ (2β − p)eiϑ](1 − eiϑ)
∣∣∣∣
=2(β − p)L√
p2 + (2β − p)2 − 2p(2β − p) cosϑ√
2(1 − cosϑ)
≥ β − p
2β.
(58)
This contradicts our condition (53). Therefore ,there is no z0 ∈ E such that |ϕ(z0)| = 1. This
implies that |ϕ(z)| < 1 (z ∈ E∗), that is,
∣∣∣∣∣∣
z∂qf(z)f(z) + p
z∂qf(z)f(z) + (2β − p)
∣∣∣∣∣∣< 1, (z ∈ E)
thus, we conclude that f ∈ MS(p, 0, β, q). 2As q → 1− we get the following result proved by Shi.et al. [13].
Corollary 2.8 If f ∈ Σp satisfies
∣∣∣∣1 +zf ′′(z)
f ′(z)− zf ′(z)
f(z)
∣∣∣∣ <β − p
2β(59)
for some real β > p, than f ∈ MS(p, 0, β).
References
[1] A.G.Alamoush and M.Darus, Some results for certain classes of spirallike meromorphic
functions, South Asian Bull.Math., 20(2014), 1-9.
[2] O.Al-Refai and M.Darus, General univalence criterion associated with the nth derivative,
Abstract and Appl. Analy., (2012), Article ID 307526, 9 pages.
[3] M.Darus and I.Faisal, A study on Beckers univalence criteria, Abstract and Appl. Analy.,(2011),
Article ID759175, 13 pages.
28 Read.S.A.Qahtan, Hamid Shamsan and S.Latha
[4] B.A.Frasin and M.Darus, On certain meromorphic functions with positive coefficients,
South Asian Bull. Math., 28(2004), 615-623.
[5] B.A.Frasin and M.Darus, Meromorphic univalent functions with positive and fixed second
coefficients, South Asian Bull. Math., 30(2006), 827 - 834.
[6] K.Hamai, T.Hayami, K.Kuroki and S.Owa, Coefficient estimates of functions in the class
concerning with spirallike functions, Appl. Math.,11(2011), 189-196.
[7] I.S.Jack, functions starlike and convex of order α, J. Lond. Math. Soc., 3(1971), 469-474.
[8] F.H.Jackson, On q-functions and a certain difference operator, Trans. Royal Soc. Edin-
burgh,46(1909), 253-281.
[9] A.Janteng, S.A.Halim and M.Darus, A new subclass of harmonic univalent functions, South
Asian Bull. Math., 31(2007), 81-88.
[10] S.Latha and L.Shivarudrappa, Partial sums of some meromorphic functions, Jour. Ineq.
Pure and Appl. Math., 7(4)(2006), Article 140.
[11] M.L.Mogra, T.R.Reddy and O.P.Juneja, Meromorphic univalent functions, Bull. Austral.
Math. Soc., 32(1985), 161-176.
[12] L.Shi, Z.G.Wang, and J.P.Yi, A new class of meromorphic functions associated with spi-
rallike functions, Jour. Appl. Math.,(2012), 494-917.
[13] L.Shi, Z.W.Wang and M.H.Zeng, Some subclasses of multivalent spirallike meromorphic
functions, J. Ineq. Appl., 2013(1), 336.
[14] Y.Sun, W.P.Kuang and Z.G.Wang, On meromorphic starlike functions of reciprocal order
α, Bull. Malay. Math. Sci. Soc., 35(2)(2012), 469-477.
[15] N.Uyanik, H.Shiraishi, S.Owa and Y.Polatoglu, Reciprocal classes of p-valently spirallike
and p-valently Robertson functions, J. Ineq. Appl., (2011), Article ID 61(2011).
International J.Math. Combin. Vol.4(2019), 29-47
On the Complexity of Some Classes of Circulant Graphs and
Chebyshev Polynomials
S. N. Daoud1,2, Muhammad Kamran Siddiqui3
1.Department of MathematicsFaculty of Science, Taibah University, Al-Madinah 41411, Saudi Arabia
2. Department of Mathematics and Computer ScienceFaculty of Science, Menoufia University, Shebin El Kom 32511, Egypt
3.Department of MathematicsCOMSATS University Islamabad, Lahore Campus, 54000, Pakistan
E-mail: [email protected], [email protected]
Abstract: Deriving closed formulae of the number of spanning trees for various graphs has
attracted the attention of a lot of researchers. In this paper we derive simple and explicit
formulas for the number of spanning trees in many classes of circulant graphs using the
properties of Chebyshev polynomials. Deriving closed formulae of the number of spanning
trees for various graphs has attracted the attention of a lot of researchers. In this paper
we derive simple and explicit formulas for the number of spanning trees in many classes of
circulant graphs using the properties of Chebyshev polynomials.
Key Words: Number of spanning trees, circulant graphs, Chebyshev polynomials.
AMS(2010): 05C78.
§1. Introduction
The number of spanning trees τ(G) in a graph G (networks) is an important invariant. We
call τ(G) the complexity of G. The evaluation of this number is not only interesting from a
mathematical (computational) perspective, but also, it is an important measure of reliability of
a network and designing electrical circuits. Some computationally hard problems such as the
travelling salesman problem can be solved approximately by using spanning trees. In this work
we consider finite undirected graph with no loops or multiple edges. Let G be such a graph of
n vertices. A spanning tree for a graph G is a subgraph of G that is a tree and contains all
vertices of τ(G). The number of spanning trees of G, is the total number of distinct spanning
subgraphs of G that are trees. A classic result of Kirchhoff [?] can be used to determine the
number τ(G) for G(V,E). Let V = {v1, v2, · · · , vn}. The Kirchhoff matrix H is defined as n×ncharacteristic matrix H = D−A, where D is the diagonal matrix of the degrees of G and A is
the adjacency matrix of G. Then the matrix H − [aij ] is defined as follows:
(1) aij , when vi and vj are adjacent and i 6= j;
(ii) aij , is equal to the degree of vertex vi if i = j;
(iii) aij = 0, otherwise. All of co-factors of H are equal to τ(G).
1Received April 25, 2019, Accepted November 23, 2019.
30 S. N. Daoud and Muhammad Kamran Siddiqui
There are more than one method for calculating τ(G). Let µ1 ≥ µ2 ≥ · · · ≥ µp denote the
eigenvalues of matrix of a p point graph. It can be easily shown that µp = 0. Kelmans and
Chelnokov [2] proved that The formula for the number of spanning trees in a d-regular graph
can be expressed as
τ(G) =1
p
[ p−1∏
k=1
(d− µk)]
where µ0 = d, µ1, µ2, · · · , µp−1 are the eigenvalues of the corresponding adjacency matrix of
the graph. Many works have conceived techniques to derive the number of spanning tree of a
graph can be found at [3-12]. The circulant graphs are an important class of graphs. Among
other applications, they are used in the design of local area networks, see [13-19].
Let 1 ≤ a1 ≤ a2 ≤ a3 ≤ · · · ≤ ak ≤ n2 , where n and ai(i = 1, 2, . . . , k) are positive integers.
An undirected circulant graph Cn(a1, a2, a3, · · · , ak) is a regular graph whose set of vertices is
V = {0, 1, 2, [dots, n−1} and whose set of edges is E = {i, i+ai(mod n)/i = 0, 1, 2, · · · , n−1, j =
1, 2, · · · , k}. If ak ≤ n2 , then Cn(a1, a2, a3, · · · , ak) is a 2k-regular graph; if ak = n
2 , then it
is a 2k − 1-regular one, see Nikolopoulos [20] and Papadopoulos [21]. The well known formula
τ(Cn(1, 2)) = nF 2n , where Fn is the nth Fibonacci number, see Kleiman, and Golden [22]. We
have obtained another proof for this formula in Theorem 3.3. The formulas of τ(C2n(1, n)),
τ(C3n(1, n)), τ(C4n(1, n)) can be found in Yuanping, et. al.[23].
§2. Chebyshev Polynomial
In this section we introduce some relations concerning Chebyshev polynomials of the first and
second kind which we use it in our computations. We begin from their definitions, see Yuanping,
et. al.[24]. Let An(x) be n× n matrix such that:
An(x) =
2x −1 0 . . . . . .
−1 2x −1. . .
...
0. . .
. . .. . . 0
.... . .
. . . 2x −1...
. . . 0 −1 2x
where all other elements are zeros.
Further, we recall that the Chebyshev polynomials of the first kind are defined by:
Tn(x) = cos(n arccos). (1)
The Chebyshev polynomials of the second kind are defined by
Un−1(x) =1
n
d
dxTn(x) =
sin(n arccos)
sin(arccos). (2)
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 31
It is easily verified that
Un(x) − 2xUn−1(x) + Un−2(x) = 0. (3)
It can then be shown from this recursion that by expanding one gets
Un(x) = det(An(x)), n ≥ 1. (4)
Furthermore by using standard methods for solving the recursion (3), one obtains the
explicit formula
Un(x) =1
2√x2 − 1
[(x+
√x2 − 1
)n+1 −(x−
√x2 − 1
)n+1], n ≥ 1, (5)
where the identity is true for all complex (except at x = ±1 , where the function can be taken
as the limit). The definition of easily yields its zeros and it can therefore be verified that
Un−1(x) = 2n−1[ n−1∏
j=1
(x− cosjπ
n)]. (6)
One further notes that
Un−1(−x) = (−1)n−1Un−1(x). (7)
These two results yield another formula for Un(x)
U2n−1(x) = 4n−1
[ n−1∏
j=1
(x2 − cos2jπ
n)]. (8)
Finally, a simple manipulation of the above formula yields the following formula (9), which
is extremely useful to us latter:
U2n−1
(√x+ 2
4
)=[ n−1∏
j=1
(x− 2 cosjπ
n)]
(9)
Furthermore, one can show that
U2n−1(x) =
1
2(1 − x2)
[1 − T2n
]=
1
2(1 − x2)
[1 − T2n(2 − 2x2)
](10)
and
Tn(x) =1
2
[(x+
√x2 − 1
)n+(x−
√x2 − 1
)n]. (11)
§3. Main Results
In our main results, i.e., Theorems 3.1 - 3.6 we use the following conclusion.
Lemma 3.1([25]) The Kirchhoff matrix of the circulant graph Cn(s1, s2, s3, · · · , sk) has n
32 S. N. Daoud and Muhammad Kamran Siddiqui
eigenvalues, namely: 0 and the value 2k−ε−s1j −· · ·−ε−skj −ε−s1j −· · ·−ε−skj with ε = e2πjn
for any j = {1, 2, · · · , n− 1}
Corollary 3.2 For the circulant graph Cn(s1, s2, s3, · · · , sk),
τ(Cn(s1, s2, s3, · · · , sk)) =1
n
n−1∏
j=1
(2k − ε−s1j − · · · − ε−skj − ε−s1j − · · · − ε−skj
)
=1
n
n−1∏
j=1
(k∑
i=1
(2 − 2 cosjsiπ
n)
) .
Proof The proof follows immediately from Lemma 3.1. 2Theorem 3.3 For the spanning trees of C12n with three jumps 1, 2n, 3n , we have:
τ(C12n(1, 2n, 3n)) =n
12
(√
7
4+
√3
4
)4n
+
(√7
4−√
3
4
)4n
− 1
2
× n
12
(√
7
4+
√3
4
)2n
+
(√7
4−√
3
4
)2n
+ 1
2
× n
12
(√
5
2+
√3
2
)2n
+
(√5
2−√
3
2
)2n
2
× n
12
[(√2 + 1
)n
+(√
2 − 1)n]2
× n
12
(√
11
4+
√7
4
)2n
+
(√11
4−√
7
4
)2n
− 1
2
Proof Let ε = e2πi12n . Applying Lemma 3.1, we have
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(6 − ε−j − ε−2nj − ε−3nj − εj − ε2nj − ε3nj
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
4πjn
12n− 2 cos
6πjn
12n
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
πj
3− 2 cos
πj
2
)
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 33
=1
12n
12n−1∏
j=1,2∤j
(6 − 2 cos
2πj
12n− 2 cos
πj
3
)
×12n−1∏
j=1,2|j
(6 − 2 cos
2πj
12n− 2 cos
πj
3− 2 cos
πj
2
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
πj
3
)
×12n−1∏
j=1
6 − 2 cos 2πj12n
− 2 cos πj3 − 2 cos πj
2
6 − 2 cos 2πj12n
− 2 cos πj3
If we put j = 2j′ in the second term for some integer j′ we get
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
πj
3
)
×6n−1∏
j=1
6 − 2 cos 2πj6n
− 2 cos 2πj3 − 2 cosπj
6 − 2 cos 2πj6n
− 2 cos πj3
=1
12n
12n−1∏
j=1,2∤j,3∤j
(5 − 2 cos
2πj
12n
) 12n−1∏
j=1,2|j,3|j
(4 − 2 cos
2πj
12n
)
×12n−1∏
j=1,2|j,3|j
(8 − 2 cos
2πj
12n
) 12n−1∏
j=1,2|j,3|j
(7 − 2 cos
2πj
12n
)
×
6n−1∏j=1,2∤j
(8 − 2 cos 2πj
6n− 2 cos 2πj
3
) 6n−1∏j=1,2∤j
(4 − 2 cos 2πj
6n− 2 cos 2πj
3
)
6n−1∏j=1
(6 − 2 cos 2πj
6n− 2 cos πj
3
) .
Thus,
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(5 − 2 cos
2πj
12n
) 2n−1∏
j=1
(5 − 2 cos 2πj
2n
) (4 − 2 cos 2πj
2n
)(8 − 2 cos 2πj
2n
) (7 − 2 cos 2πj
2n
)
×6n−1∏
j=1
8 − 2 cos 2πj6n
5 − 2 cos 2πj6n
4n−1∏
j=1
7 − 2 cos 2πj4n
5 − 2 cos 2πj4n
×
6n−1∏j=1
(8 − 2 cos 2πj
6n− 2 cos 2πj
3
) 3n−1∏j=1
(4 − 2 cos 2πj
3n− 2 cos 4πj
3
)
6n−1∏j=1
(6 − 2 cos 2πj
6n− 2 cos πj
3
) 3n−1∏j=1
(8 − 2 cos 2πj
3n− 2 cos 4πj
3
) .
34 S. N. Daoud and Muhammad Kamran Siddiqui
So, we get that
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(5 − 2 cos
2πj
12n
) 2n−1∏
j=1
(5 − 2 cos 2πj
2n
) (4 − 2 cos 2πj
2n
)(8 − 2 cos 2πj
2n
) (7 − 2 cos 2πj
2n
)
×6n−1∏
j=1
8 − 2 cos 2πj6n
5 − 2 cos 2πj6n
4n−1∏
j=1
7 − 2 cos 2πj4n
5 − 2 cos 2πj4n
×
6n−1∏j=1,3|j
(9 − 2 cos 2πj
6n
) 6n−1∏j=1,3|j
(6 − 2 cos 2πj
6n
)
6n−1∏j=1,3|j
(7 − 2 cos 2πj
6n
) 6n−1∏j=1,3|j
(4 − 2 cos 2πj
6n
)
×
3n−1∏j=1
(6 − 2 cos 2πj
3n− 4 cos2 2πj
3
)
3n−1∏j=1
(10 − 2 cos 2πj
3n− 4 cos2 2πj
3
) ,
i.e.,
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(5 − 2 cos
2πj
12n
) 2n−1∏
j=1
(5 − 2 cos 2πj
2n
) (4 − 2 cos 2πj
2n
)(8 − 2 cos 2πj
2n
) (7 − 2 cos 2πj
2n
)
×6n−1∏
j=1
8 − 2 cos 2πj6n
5 − 2 cos 2πj6n
4n−1∏
j=1
7 − 2 cos 2πj4n
5 − 2 cos 2πj4n
×
6n−1∏j=1
(9 − 2 cos 2πj
6n
) 6n−1∏j=1
(6−2 cos 2πj6n )
(9−2 cos 2πj6n )
6n−1∏j=1
(7 − 2 cos 2πj
6n
) 6n−1∏j=1
(4−2 cos 2πj6n )
(7−2 cos 2πj6n )
×
3n−1∏j=1,3|j
(5 − 2 cos 2πj
3n
) 3n−1∏j=1,3|j
(2 − 2 cos 2πj
3n
)
3n−1∏j=1,3|j
(9 − 2 cos 2πj
3n
) 3n−1∏j=1,3|j
(6 − 2 cos 2πj
3n
)
Furthermore,
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(5 − 2 cos
2πj
12n
) 2n−1∏
j=1
(5 − 2 cos 2πj
2n
) (4 − 2 cos 2πj
2n
)(8 − 2 cos 2πj
2n
) (7 − 2 cos 2πj
2n
)
×6n−1∏
j=1
8 − 2 cos 2πj6n
5 − 2 cos 2πj6n
4n−1∏
j=1
7 − 2 cos 2πj4n
5 − 2 cos 2πj4n
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 35
×
6n−1∏j=1
(9 − 2 cos 2πj
6n
) 2n−1∏j=1
(6−2 cos 2πj6n )
(9−2 cos 2πj6n )
6n−1∏j=1
(7 − 2 cos 2πj
6n
) 2n−1∏j=1
(4−2 cos 2πj6n )
(7−2 cos 2πj6n )
×
3n−1∏j=1
(5 − 2 cos 2πj
3n
) 3n−1∏j=1
(2−2 cos 2πj3n )
(5−2 cos 2πj3n )
3n−1∏j=1
(9 − 2 cos 2πj
3n
) 3n−1∏j=1
(6−2 cos 2πj3n )
(9−2 cos 2πj3n )
.
We get that
τ(C12n(n, 2n, 3n)) =1
12n
12n−1∏
j=1
(5 − 2 cos
2πj
12n
) 2n−1∏
j=1
(5 − 2 cos 2πj
2n
) (4 − 2 cos 2πj
2n
)(8 − 2 cos 2πj
2n
) (7 − 2 cos 2πj
2n
)
×6n−1∏
j=1
8 − 2 cos 2πj6n
5 − 2 cos 2πj6n
4n−1∏
j=1
7 − 2 cos 2πj4n
5 − 2 cos 2πj4n
×
6n−1∏j=1
(9 − 2 cos 2πj
6n
) 2n−1∏j=1
(6−2 cos 2πj6n )
(9−2 cos 2πj6n )
6n−1∏j=1
(7 − 2 cos 2πj
6n
) 2n−1∏j=1
(4−2 cos 2πj6n )
(7−2 cos 2πj6n )
×
3n−1∏j=1
(5 − 2 cos 2πj
3n
) n−1∏j=1
(2−2 cos 2πjn )
(5−2 cos 2πjn )
3n−1∏j=1
(9 − 2 cos 2πj
n
) n−1∏j=1
(6−2 cos 2πjn )
(9−2 cos 2πjn )
.
Thus,
τ(C12n(n, 2n, 3n)) =1
12n× U2
12n−1
(√7
4
)×U2
2n−1
(√74
)× U2
2n−1
(√32
)
U22n−1
(32
)× U2
2n−1
(√52
)
×U2
4n−1
(32
)× U2
6n−1
(√52
)
U26n−1
(√72
)× U2
6n−1
(√72
)
×U2
6n−1
(√114
)× U2
2n−1
(32
)× U2
2n−1
(√2)× U2
3n−1
(74
)× U2
n−1
(114
)n2
U26n−1
(32
)× U2
2n−1
(√32
)× U2
n−1
(√74
)× U2
3n−1
(√114
)× U2
n−1
(√2) ,
which implies that
τ(C12n(n, 2n, 3n)) =n
12
(√
7
4+
√3
4
)4n
+
(√7
4−√
3
4
)4n
− 1
2
36 S. N. Daoud and Muhammad Kamran Siddiqui
× n
12
(√
7
4+
√3
4
)2n
+
(√7
4−√
3
4
)2n
+ 1
2
× n
12
[(√5
2+
√3
2
)2n
+(√5
2−√
3
2
)2n]2
× n
12
[(√2 + 1
)n
+(√
2 − 1)n]2
× n
12
(√
11
4+
√7
4
)2n
+
(√11
4−√
7
4
)2n
− 1
2
,
where (6), (8) , (9) and (10) are used to derive the last two steps. 2Theorem 3.4 For the spanning trees of C6n with three jumps 1, 3n, 6n, we have
τ(C12n(1, 3n, 6n)) =3n
4
(√
5
2+
√3
2
)6n
+
(√5
2−√
3
2
)6n
2
×3n
4
[(√2 + 1
)3n
+(√
2 − 1)3n]2
Proof Let ε = e2πi12n . Applying Lemma 3.1, we have
τ(C12n(1, 3n, 6n)) =1
12n
12n−1∏
j=1
(6 − ε−j − ε−3nj − ε−6nj − εj − ε3nj − ε6nj
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
6πjn
12n− 2 cos
12πjn
12n
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
πj
2− 2 cosπj
)
=1
12n
12n−1∏
j=1,2∤j
(8 − 2 cos
2πj
12n
) 12n−1∏
j=1,2|j
(4 − 2 cos
2πj
12n− 2 cos
πj
2
).
Whence,
τ (C12n(1, 3n, 6n)) =1
12n
12n−1∏
j=1
(8 − 2 cos
2πj
12n
) 6n−1∏
j=1
4 − 2 cos 2πj
12n− 2 cos πj
2
8 − 2 cos 2πj
6n
=1
12n
12n−1∏j=1
(8 − 2 cos 2πj
12n
)
6n−1∏j=1
(8 − 2 cos 2πj
6n
) ×6n−1∏
j=1,2∤j
(6 − 2 cos
2πj
6n
) 6n−1∏
j=1,2|j
(6 − 2 cos
2πj
6n
).
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 37
Thus,
τ(C12n(1, 3n, 6n)) =1
12n
12n−1∏j=1
(8 − 2 cos 2πj
12n
)
6n−1∏j=1
(8 − 2 cos 2πj
6n
)
6n−1∏
j=1
(6 − 2 cos
2πj
6n
)
×6n−1∏
j=1
2 − 2 cos 2πj6n
6 − 2 cos 2πj6n
=1
12n
12n−1∏j=1
(8 − 2 cos 2πj
12n
)
6n−1∏j=1
(8 − 2 cos 2πj
6n
)
6n−1∏
j=1
(6 − 2 cos
2πj
6n
)
×3n−1∏
j=1
2 − 2 cos 2πj3n
6 − 2 cos 2πj3n
,
which implies that,
τ(C12n(1, 3n, 6n)) =1
12n×U2
12n−1
(√52
)× (3n)2 × U2
6n−1(√
2)
U26n−1
(√52
)× U2
3n−1(√
2)
=3n
4×U2
12n−1
(√52
)× U2
6n−1(√
2)
U26n−1
(√52
)× U2
3n−1(√
2),
i.e.,
τ(C12n(1, 3n, 6n)) =3n
4
(√
5
2+
√3
2
)6n
+
(√5
2−√
3
2
)6n
2
×3n
4
[(√2 + 1
)3n
+(√
2 − 1)3n]2,
where (6), (8) , (9) and (10) are used to derive the last two steps. 2Theorem 3.5 For the spanning trees of C12n with three jumps 1, 3n, 4n , we have:
τ(C12n(1, 3n, 4n)) =n
12
(√
9
4+
√5
4
)4n
+
(√9
4−√
5
4
)4n
− 1
2
×
(√
7
4+
√3
4
)2n
+
(√7
4−√
3
4
)2n
+ 1
2
38 S. N. Daoud and Muhammad Kamran Siddiqui
×
(√
3
2+
√1
2
)2n
+
(√3
2−√
1
2
)2n
2
×[(√
2 + 1)n
+(√
2 − 1)n]2
×
(√
11
4+
√7
4
)2n
+
(√11
4−√
7
4
)2n
− 1
2
Proof Let ε = e2πi12n . Applying Lemma 3.1, we have
τ(C12n(1, 3n, 4n)) =1
12n
12n−1∏
j=1
(6 − ε−j − ε−3nj − ε−4nj − εj − ε3nj − ε4nj
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
πj
2− 2 cos
2πj
3
)
=1
12n
12n−1∏
j=1,2∤j
(6 − 2 cos
2πj
12n− 2 cos
πj
3
)
×12n−1∏
j=1,2|j
(6 − 2 cos
2πj
12n− 2 cos
πj
3− 2 cos
πj
2
)
=1
12n
12n−1∏
j=1
(6 − 2 cos
2πj
12n− 2 cos
πj
3
)
×6n−1∏
j=1
6 − 2 cos 2πj6n
− 2 cosπj − 2 cos 4πj3
6 − 2 cos 2πj6n
− 2 cos 4πj3
=1
12n
12n−1∏
j=1,3∤j
(7 − 2 cos
2πj
12n
) 12n−1∏
j=1,3|j
(4 − 2 cos
2πj
12n
)
×6n−1∏
j=1
8 − 2 cos 2πj6n
− 2 cosπj − 2 cos2 2πj3
6n−1∏j=1
8 − 2 cos 2πj6n
− 2 cos2 2πj3
.
Thus,
τ (C12n(1, 3n, 4n)) =1
12n
12n−1∏
j=1
(7 − 2 cos
2πj
12n
) 4n−1∏
j=1
4 − 2 cos 2πj
4n
7 − 2 cos 2πj
4n
×
6n−1∏j=1,2∤j
(10 − 2 cos 2πj
6n− 4 cos2 2πj
3
) 6n−1∏j=1,2|j
(6 − 2 cos 2πj
6n− 4 cos2 2πj
3
)
6n−1∏j=1,3∤j
(7 − 2 cos 2πj
6n
) 6n−1∏j=1,3|j
(4 − 2 cos 2πj
6n
) .
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 39
We therefore get that
τ(C12n(1, 3n, 4n)) =1
12n
12n−1∏
j=1
(7 − 2 cos
2πj
12n
) 4n−1∏
j=1
4 − 2 cos 2πj4n
7 − 2 cos 2πj4n
×
6n−1∏j=1
(10 − 2 cos 2πj
6n− 4 cos2 2πj
3
) 3n−1∏j=1
6−2 cos 2πj3n
−4 cos2 2πj3
10−2 cos 2πj3n
−4 cos2 2πj3
6n−1∏j=1
(7 − 2 cos 2πj
6n
) 2n−1∏j=1
4−2 cos 2πj2n
7−2 cos 2πj2n
,
i.e.,
τ(C12n(1, 3n, 4n)) =1
12n
12n−1∏
j=1
(7 − 2 cos
2πj
12n
) 4n−1∏
j=1
4 − 2 cos 2πj4n
7 − 2 cos 2πj4n
×
6n−1∏j=1,3∤j
(9 − 2 cos 2πj
6n
) 6n−1∏j=1,3|j
(6 − 2 cos 2πj
6n
)
6n−1∏j=1
(7 − 2 cos 2πj
6n
) 2n−1∏j=1
4−2 cos 2πj2n
7−2 cos 2πj2n
×
6n−1∏j=1,3∤j
(5 − 2 cos 2πj
3n
) 6n−1∏j=1,3|j
(7 − 2 cos 2πj
3n
)
6n−1∏j=1,3∤j
(9 − 2 cos 2πj
3n
) 6n−1∏j=1,3|j
(6 − 2 cos 2πj
3n
) ,
which implies that
τ(C12n(1, 3n, 4n)) =1
12n
12n−1∏
j=1
(7 − 2 cos
2πj
12n
) 4n−1∏
j=1
4 − 2 cos 2πj4n
7 − 2 cos 2πj4n
×
6n−1∏j=1
(9 − 2 cos 2πj
6n
) 6n−1∏j=1
6−2 cos 2πj6n
9−2 cos 2πj6n
6n−1∏j=1
(7 − 2 cos 2πj
6n
) 2n−1∏j=1
4−2 cos 2πj2n
7−2 cos 2πj2n
×
3n−1∏j=1
(5 − 2 cos 2πj
3n
) 3n−1∏j=1
7−2 cos 2πj3n
5−2 cos 2πj3n
3n−1∏j=1
(9 − 2 cos 2πj
3n
) n−1∏j=1
6−2 cos 2πjn
9−2 cos 2πjn
.
Thus,
τ(C12n(1, 3n, 4n)) =1
12n×
U24n−1
(√32
)×U2
12n−1
(32
)
U24n−1
(√32
)
U22n−1
(√32
)×U2
6n−1
(32
)
U22n−1
(√32
)
40 S. N. Daoud and Muhammad Kamran Siddiqui
×
U26n−1
(√114
)×U2
2n−1
(√
2
)
U22n−1
(√114
) ×n2×U2
3n−1
(√74
)
U23n−1
(√114
)
U2n−1
(√
2
)×U2
3n−1
(114
)
U2n−1
(√114
)
.
We have
τ(C12n(1, 3n, 4n)) =n
12
(√
9
4+
√5
4
)4n
+
(√9
4−√
5
4
)4n
− 1
2
×
(√
7
4+
√3
4
)2n
+
(√7
4−√
3
4
)2n
+ 1
2
×
(√
3
2+
√1
2
)2n
+
(√3
2−√
1
2
)2n
2
×[(√
2 + 1)n
+(√
2 − 1)n]2
×
(√
11
4+
√7
4
)2n
+
(√11
4−√
7
4
)2n
− 1
2
,
where (6), (8) , (9) and (10) are used to derive the last two steps. 2Theorem 3.6 For the spanning trees of C12n with three jumps 1, 2n, 3n, 6n , we have
τ (C12n(1, 2n, 3n, 6n)) =n
12
[(√11
4+
√7
4
)4n
+
(√11
4−√
7
4
)4n
− 1
]2
×[(√
11
4+
√7
4
)2n
+
(√11
4−√
7
4
)2n
+ 1
]2
×[(√
13
4+
√9
4
)4n
+
(√13
4−√
9
4
)4n
+ 1
]2
×[(√
7
2+
√5
2
)2n
+
(√7
2−√
5
2
)2n]2
×[(√
2 + 1)n
+(√
2 − 1)n]2
×[(√
7
4+
√3
4
)2n
+
(√7
4−√
3
4
)2n
+ 1
]2
.
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 41
Proof Let ε = e2πi12n . Applying Lemma 3.1, we get the required result. 2
Theorem 3.7 For the spanning trees of C12n with four jumps 1, 2n, 3n, 4n, we have
τ(C12n(1, 2n, 3n, 4n)) =n
12
(√
5
2+
√3
2
)6n
+
(√5
2−√
3
2
)6n
2
×
(√
5
2+
√3
2
)2n
+
(√5
2−√
3
2
)2n
+ 1
2
×
(√
7
2+
√5
2
)2n
+
(√7
2−√
5
2
)2n
− 1
2
×[(√
2 + 1)n
+(√
2 − 1)n]2
.
Proof Let ε = e2πi12n . Applying Lemma 3.1, we have
τ(C12n(1, 2n, 3n, 4n))
=1
12n
12n−1∏
j=1
(8 − ε−j − ε−2nj − ε−3nj − ε−4nj − εj − ε2nj − ε3nj − ε4nj
)
=1
12n
12n−1∏
j=1
(8 − 2 cos
2πjn
12n− 2 cos
6πjn
12n− 2 cos
8πjn
12n
)
=1
12n
12n−1∏
j=1
(8 − 2 cos
πj
3− 2 cos
πj
2− 2 cos
2πj
3
)
=1
12n
12n−1∏
j=1,2∤j
(8 − 2 cos
πj
3− 2 cos
πj
2− 2 cos
2πj
3
)
×12n−1∏
j=1,2|j
(8 − 2 cos
πj
3− 2 cos
πj
2− 2 cos
2πj
3
)
=1
12n
12n−1∏
j=1
(8 − 2 cos
πj
3− 2 cos
πj
2− 2 cos
2πj
3
)
×6n−1∏
j=1
8 − 2 cos πj3 − 2 cos πj
2 − 2 cos 2πj3
8 − 2 cos 2πj6n
− 2 cos 2πj3 − 2 cos 4πj
3
.
Thus,
τ(C12n(1, 2n, 3n, 4n)) =1
12n
12n−1∏
j=1,3∤j
(9 − 2 cos
2πj
12n− 2 cos
πj
3
)
42 S. N. Daoud and Muhammad Kamran Siddiqui
×12n−1∏
j=1,3|j
(6 − 2 cos
2πj
12n− 2 cos
πj
3
)
×
6n−1∏j=1
(10 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3 − 2 cosπj
)
6n−1∏j=1
(10 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
) ,
i.e.,
τ(C12n(1, 2n, 3n, 4n)) =1
12n
12n−1∏
j=
(9 − 2 cos
2πj
12n− 2 cos
πj
3
)
×4n−1∏
j=1
6 − 2 cos 2πj4n
− 2 cos πj3
9 − 2 cos 2πj4n
− 2 cos πj3
×
6n−1∏j=1
(10 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3 − 2 cosπj
)
6n−1∏j=1
(10 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
) .
We get that
τ(C12n(1, 2n, 3n, 4n)) =1
12n
12n−1∏
j=1
(9 − 2 cos
2πj
12n− 2 cos
πj
3
)
×4n−1∏
j=1
6 − 2 cos 2πj4n
− 2 cos πj3
9 − 2 cos 2πj4n
− 2 cos πj3
×
6n−1∏j=1,2∤j
(12 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
)
6n−1∏j=1,3∤j
(10 − 2 cos 2πj
6n
)
×
6n−1∏j=1,2|j
(8 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
)
6n−1∏j=1,3|j
(4 − 2 cos 2πj
6n
) ,
which implies that
τ(C12n(1, 2n, 3n, 4n)) =1
12n
12n−1∏
j=1,2∤j,3∤j
(8 − 2 cos
2πj
12n
) 12n−1∏
j=1,2|j,3|j
(7 − 2 cos
2πj
12n
)
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 43
×12n−1∏
j=1,2∤j,3|j
(11 − 2 cos
2πj
12n
) 12n−1∏
j=1,2∤j,3|j
(10 − 2 cos
2πj
12n
)
×
4n−1∏j=1,2∤j
(8 − 2 cos 2πj
4n
) 4n−1∏j=1,2|j
(4 − 2 cos 2πj
4n
)
4n−1∏j=1,2∤j
(11 − 2 cos 2πj
4n
) 4n−1∏j=1,2|j
(7 − 2 cos 2πj
4n
)
×
6n−1∏j=1
(12 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
)
6n−1∏j=1
(10 − 2 cos 2πj
6n
)
×
6n−1∏j=1
8−2 cos 2πj6n
−2 cos 2πj3 −4 cos2 2πj
3
12−2 cos 2πj6n
−2 cos 2πj3 −4 cos2 2πj
3
6n−1∏j=1
4−2 cos 2πj6n
10−2 cos 2πj6n
.
Thus,
τ(C12n(1, 2n, 3n, 4n)) =1
12n
12n−1∏
j=1
(8 − 2 cos
2πj
12n
)
×12n−1∏
j=1
(8 − 2 cos 2πj
12n
) (7 − 2 cos 2πj
12n
)(11 − 2 cos 2πj
12n
) (10 − 2 cos 2πj
12n
)
×12n−1∏
j=1
10 − 2 cos 2πj12n
]
8 − 2 cos 2πj12n
×12n−1∏
j=1
11 − 2 cos 2πj12n
8 − 2 cos 2πj12n
×
4n−1∏j=1
(8 − 2 cos 2πj
4n
) 4n−1∏j=1
4−2 cos 2πj4n
8−2 cos 2πj4n
4n−1∏j=1
(11 − 2 cos 2πj
4n
) 4n−1∏j=1
7−2 cos 2πj4n
11−2 cos 2πj4n
×
6n−1∏j=1
(12 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
)
6n−1∏j=1
(10 − 2 cos 2πj
6n
)
×
6n−1∏j=1
8−2 cos 2πj6n
−2 cos 2πj3 −4 cos2 2πj
3
12−2 cos 2πj6n
−2 cos 2πj3 −4 cos2 2πj
3
6n−1∏j=1
4−2 cos 2πj6n
10−2 cos 2πj6n
.
We get that
44 S. N. Daoud and Muhammad Kamran Siddiqui
τ(C12n(1, 2n, 3n, 4n)) =1
12n
12n−1∏
j=1
(8 − 2 cos
2πj
12n
)
×12n−1∏
j=1
(8 − 2 cos 2πj
12n
) (7 − 2 cos 2πj
12n
)(11 − 2 cos 2πj
12n
) (10 − 2 cos 2πj
12n
)
×12n−1∏
j=1
10 − 2 cos 2πj12n
8 − 2 cos 2πj12n
12n−1∏
j=1
11 − 2 cos 2πj12n
8 − 2 cos 2πj12n
×
4n−1∏j=1
(8 − 2 cos 2πj
4n
) 4n−1∏j=1
4−2 cos 2πj4n
8−2 cos 2πj4n
4n−1∏j=1
(11 − 2 cos 2πj
4n
) 4n−1∏j=1
7−2 cos 2πj4n
11−2 cos 2πj4n
×
6n−1∏j=1
(12 − 2 cos 2πj
6n− 2 cos 2πj
3 − 4 cos2 2πj3
)
6n−1∏j=1
(10 − 2 cos 2πj
6n
)
×
2n−1∏j=1
8−2 cos 2πj2n
−2 cos 2πj3 −4 cos2 2πj
3
12−2 cos 2πj2n
−2 cos 2πj3 −4 cos2 2πj
3
3n−1∏j=1
4−2 cos 2πj3n
10−2 cos 2πj3n
,
i.e.,
τ(C12n(1, 2n, 3n, 4n)) =1
12n
×U2
12n−1
(√114
)× U2
2n−1
(√114
)× U2
2n−1
(√72
)× U2
4n−1
(√72
)× U2
6n−1
(√134
)
U22n−1
(√72
)× U2
2n−1
(√134
)× U2
4n−1
(√114
)× U2
6n−1
(√114
)
×U2
6n−1
(√114
)× U2
2n−1
(√2)× U2
3n−1
(√74
)× U2
n−1
(√114
)× U2
2n−1
(√114
)
U22n−1
(√114
)× U2
n−1
(√74
)× U2
3n−1
(√114
)× U2
n−1
(√2)× U2
6n−1
(√114
)
So, we have
τ(C12n(1, 2n, 3n, 4n)) =n
12
(√
5
2+
√3
2
)6n
+
(√5
2−√
3
2
)6n
2
×
(√
5
2+
√3
2
)2n
+
(√5
2−√
3
2
)2n
+ 1
2
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 45
×
(√
7
2+
√5
2
)2n
+
(√7
2−√
5
2
)2n
− 1
2
×[(√
2 + 1)n
+(√
2 − 1)n]2
,
where (6), (8), (9) and (10) are used to derive the last two steps. 2Theorem 3.8 For the spanning trees of C6n with four jumps 1, 3n, 4n, 6n, we have
τ(C12n(1, 3n, 4n, 6n)) =n
12
(√
11
4+
√7
4
)8n
+
(√11
4−√
7
4
)8n
+ 1
2
×
(√
11
4+
√7
4
)2n
+
(√11
4−√
7
4
)2n
− 1
2
×
(√
3
2+
√5
2
)2n
+
(√3
2−√
5
2
)2n
2
×[(√
2 + 1)n
+(√
2 − 1)n]2
×
(√
7
4+
√3
4
)2n
+
(√7
4−√
3
4
)2n
+ 1
2
×
(√
13
4+
√9
4
)2n
+
(√13
4−√
9
4
)2n
+ 1
2
Proof Let ε = e2πi12n . Apply Lemma 3.1, we get the required result. 2
Theorem 3.9 For the spanning trees of C6n with four jumps 1, 2n, 3n, 4n, 6n, we have
τ(C12n(1, 2n, 3n, 4n, 6n)) =n
12×[(√
2 + 1)n
+(√
2 − 1)n]2
×[(√
2 +√
3)n
+(√
2 −√
3)n
− 1]2
×
(√
5
2+
√3
2
)2n
+
(√5
2−√
3
2
)2n
+ 1
2
×
(√
7
2+
√5
2
)2n
+
(√7
2−√
5
2
)2n
− 1
2
×
(√
7
2+
√5
2
)2n
+
(√7
2−√
5
2
)2n
2
46 S. N. Daoud and Muhammad Kamran Siddiqui
×
(√
7
2+
√5
2
)4n
+
(√7
2−√
5
2
)4n
+ 1
2
Proof Let ε = e2πi12n . Applying Lemma 3.1, We get the required result. 2
§4. Conclusions
The number of spanning trees in graphs (networks) is an important invariant. The evaluation
of this number is not only interesting from a mathematical (computational) perspective, but
also, it is an important measure of reliability of a network and designing electrical circuits.
Some computationally hard problems such as the travelling salesman problem can be solved
approximately by using spanning trees. Due to the high dependence of the network design and
reliability on the graph theory we prove our results in Section 3.
References
[1] G. G. Kirchhoff, Uber die Auflosung der Gleichungen, auf welche man be ider Untersuchung
der Linearen Verteilung galvanischer Storme gefuhrt wird, Ann. Phys. Chem., 72 (1847),
497-508.
[2] A. K. Kelmans and V. M. Chelnokov, A certain polynomials of a graph and graphs with
an extermal number of trees, J. Comb. Theory (B), 16 (1974), 197-214.
[3] S. N. Daoud, The deletion- contraction method for counting the number of spanning trees
of graphs, European Journal of Physical Plus, Vol.130, 10(2015), 1-14.
[4] S. N. Daoud, Number of spanning trees in different product of complete and complete
tripartite graphs, Ars Combinatoria, Vol.139 (2018), 85-103.
[5] S. N. Daoud, Complexity of graphs generated by wheel graph and their asymptotic limits,
Journal of the Egyptian Math. Soc., Vol.25, 4(2017), 424-433.
[6] Jia-Bao Liu and S. N. Daoud, Complexity of some of pyramid graphs created from a gear
graph, Symmetry, 2018, 10, 689; doi:10.3390/sym10120689.
[7] S. N. Daoud, Complexity of some families of cycle related graphs, Journal of Taibah Uni-
versity for Science, 11 (2017), 205-228.
[8] S. N. Daoud, Complexity of some special named graphs and Chebyshev polynomials, In-
ternational Journal Applied Mathematical and Statistics, Ceser Pub. Vol.32, 2(2013), 77-
84.
[9] S. N. Daoud, Generating formulas of the number of spanning trees of some special graphs,
Eur. Phys. J. Plus, Vol. 129 (2014), 1-14.
[10] S. N. Daoud, Number of spanning trees of different products of complete and complete
bipartite graphs, Journal of Mathematical Problems in Engineering, Hindawi Publishing
Corporation Volume 2014, Article ID 965105, 23 pages.
[11] S. N. Daoud, On a class of some pyramid graphs and Chebyshev polynomials, Journal
of Mathematical Problems in Engineering, Hindawi Publishing Corporation Volume 2013,
On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials 47
Article ID 820549, 11 pages.
[12] Jia-Bao Liu and S. N. Daoud, Number of spanning trees in the sequence of some graphs,
Complexity, Hindawi Publ. Corp. Volume 2019, Article ID 4271783, 22. pages 2019.
[13] A. K. Kelmans, Connectivity of probabilistic networks, Auto Remote Contr., 3(1967), 444-
460.
[14] R. S. Wilkov, Analysis and design of reliable computer networks, IEEE Trans. Commun.
Syst., CS11 (1963), 230-247.
[15] J. Louis, Asymptotics for the number of spanning trees in circulant graphs and degenerating
d-dimensional discrete tori, Annals of Combinatorics, 19(3) (2015), 513-543.
[16] Chen Xiebin, The numbers of spanning trees in undirected circulant graphs, J. Zhangzhou
Teachers College, 13(4) (2000), 1-6.
[17] Chen Xiebin, Qiuying Lin, Fuji Zhang, The number of spanning trees in odd valent circulant
graphs, Discrete Math., 282(1) (2004) 69-79.
[18] X. Yong, Talip and Acenjian, The number of spanning trees of the cubic cycle C3n and
the quadruple cycle C4n, Discrete Math., 169 (1997), 293-298.
[19] Zhang Yuanping, Yong Xuerong, M.J.Golin, The number of spanning trees in circulant
graphs, Discrete. Math., 223(1) (2000) 337-350.
[20] D. Nikolopoulos and C. Papadopoulos, The number of spanning trees in complements of
quasi-threshold graphs, J. Graphs and Combin., 20 (2004), 383-397.
[21] D. Nikolopoulos and C. Papadopoulos, On the number of spanning trees of multi-star
related graph, Inform. Process. Lett., 65 (1998), 183-188.
[22] D. J. Kleiman and B. Golden, Counting trees in a certain class of graphs, Amer Math.
Mon., 82 (1975), 40-44.
[23] Z.Yuanping, Y. Xuerong and J. Mordecai, Chebyshev polynomials and spanning trees
formulas for circulant and related graphs, Discrete Math., 298 (2005), 334-364.
[24] Z.Yuanping, Y.Xuerong and J. Mordecai, The number of spanning trees in circulant graphs,
Discrete Math., 223 (2000), 337-350.
[25] N. Biggs, Algebraic Graph Theory (Second ed.), Cambridge University Press, London,
(1993).
International J.Math. Combin. Vol.4(2019), 48-60
Some Fixed Point Results for
Multivalued Nonexpansive Mappings in Banach Spaces
G.S.Saluja
(Department of Mathematics, Govt. Kaktiya P. G. College, Jagdalpur - 494001, Chhattisgarh, India)
E-mail: [email protected]
Abstract: In this paper, we deal with the approximation of fixed point for multivalued
nonexpansive mappings via a new three-step algorithm which is independent and faster than
the iterative process discussed by Agarwal et al. [3] and establish some strong convergence
theorems and a weak convergence theorem in the setting of Banach spaces. Our results
extend and generalize the previous works from the existing literature.
Key Words: Multivalued nonexpansive mapping, three-step iteration scheme, fixed point,
weak and strong convergence, Banach space.
AMS(2010): 47H10, 54H25.
§1. Introduction
Throughout this paper, let X be a real Banach space with the norm ‖.‖. Let N denotes the set
of all positive integers and let F (T ) denotes the set of all fixed points of the mapping T .
Let K be a subset of X . A subset K is called proximal if for each x ∈ X , there exists
an element k ∈ K such that d(x, k) = inf{‖x − y‖ : y ∈ K} = d(x,K). It is well known that
a weakly compact convex subset of a Banach space and closed convex subsets of a uniformly
convex Banach space are Proximal.
We shall denote CB(K), C(K) and P (K) by the families of all nonempty closed and
bounded subsets, nonempty compact subsets and nonempty proximal subsets of K, respec-
tively. Let H denote the Hausdorff metric induced by the metric d of X , that is,
H(A,B) = max{supx∈A
d(x,B), supy∈B
d(y,A)}
for every A,B ∈ CB(X ), where d(x,B) = inf{‖x− y‖ : y ∈ B}.A multivalued mapping T : K → CB(K) is said to be a contraction if there exists a constant
b ∈ [0, 1) such that for any x, y ∈ K,
H(T x, T y) ≤ b ‖x− y‖,
1Received May 30, 2019, Accepted November 25, 2019.
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces 49
and T is said to be nonexpansive if
H(T x, T y) ≤ ‖x− y‖,
for all x, y ∈ K. A point x ∈ K is called a fixed point of T if x ∈ T x.
Later, an interesting and rich fixed point theory for such maps was developed which has
applications in control theory, convex optimization, differential inclusion and economics (see
[5] and references cited therein). Moreover, the existence of fixed points for multivalued non-
expansive mappings in uniformly convex Banach spaces was proved by Lim [7]. Many authors
have studied the fixed point for multivalued mappings (e.g., see [4, 6, 8, 10, 15, 16, 19]).
In 2005, Sastry and Babu [11] obtained the convergence results from single valued mappings
to multivalued mappings by defining Ishikawa and Mann iterates for multivalued mappings with
a fixed point. They considered the following:
Let K be a nonempty convex subset of X , T : K → P (K) is a multivalued mapping with
p ∈ F (T ).
(i) The Mann iteration is defined by
{x1 = x ∈ K,
xn+1 = (1 − αn)xn + αnsn, n ∈ N,(1.1)
where {αn} is a real sequence in (0, 1) and sn ∈ T xn such that ‖sn − p‖ = d(p, T xn).
(ii) The Ishikawa iteration is defined by
x1 = x ∈ K,xn+1 = (1 − αn)xn + αnrn,
yn = (1 − βn)xn + βnsn, n ∈ N,
(1.2)
where {αn} and {βn} are real sequences in (0, 1), ‖sn − rn‖ = d(T xn, T yn) and ‖rn − p‖ =
d(T yn, T p) for sn ∈ T xn and rn ∈ T yn. They established some strong and weak convergence
results of the above iterates for multivalued nonexpansive mappings T under some appropriate
conditions.
In 2007, Panyanak [10] extended the results of Sastry and Babu [11] to a uniformly convex
Banach space and also modified the above Ishikawa iterative scheme as follows:
Let T : K → P (K) be a multi-valued mapping.
{x1 = x ∈ K,yn = (1 − βn)xn + βnzn, n ∈ N,
(1.3)
where {βn} is a real sequence in [0, 1], zn ∈ T xn and un ∈ T xn are such that ‖zn − un‖ =
50 G.S.Saluja
d(un, T xn) and ‖xn − un‖ = d(xn, F (T )), respectively, and
{x1 = x ∈ K,
xn+1 = (1 − αn)xn + αnz′n, n ∈ N,
(1.4)
where {αn} is a real sequence in [0, 1], z′n ∈ T xn and vn ∈ T xn are such that ‖z′n − vn‖ =
d(vn, T xn) and ‖yn − vn‖ = d(yn, F (T )), respectively and proved a convergence theorem of
Mann iterates for a mapping defined on a noncompact domain. Later in 2008, Song and
Wang [14] proved strong convergence theorems of Mann and Ishikawa iterates for multivalued
nonexpansive mappings under some appropriate control conditions. Furthermore, they also
gave an affirmative answer to Panyanak’s open question in [10].
Recently, Abbas and Nazir [1] introduced and studied the following iteration scheme: let
K be a nonempty subset of a Banach space X and T be a nonlinear mapping of K into itself.
Then the sequence {xn} in K is defined by
x1 = x ∈ K,xn+1 = (1 − αn)T yn + αnT zn,
yn = (1 − βn)T xn + βnT zn,
zn = (1 − γn)xn + γnT xn, n ∈ N,
(1.5)
where {αn}, {βn} and {γn} are real sequences in (0, 1). They showed that this process converges
faster than both Picard and the Agarwal et al. ([3]) and in support gave analytic proof by a
numerical example (for more details, see [1]).
Motivated by Sastry and Babu [11], Panyanak [10] and Song and Wang [14], we first give
a multivalued version of the iteration scheme (1.5) of Abbas and Nazir [1] and then study its
convergence analysis in the setting of Banach spaces. We define our iteration scheme as follows:
x1 = x ∈ K,xn+1 = (1 − αn)vn + αnwn,
yn = (1 − βn)un + βnwn,
zn = (1 − γn)xn + γnun, n ∈ N,
(1.6)
where {αn}, {βn} and {γn} are real sequences in (0, 1), un ∈ T xn, vn ∈ T yn and wn ∈ T zn
such that ‖wn − un‖ = d(T zn, T xn), ‖vn − wn‖ = d(T yn, T zn), ‖vn − un‖ = d(T yn, T xn),
‖un+1 − vn‖ = d(T xn+1, T yn) and ‖un+1 − wn‖ = d(T xn+1, T zn), respectively.
Now, we recall the following definitions.
Definition 1.1 A Banach space X is said to satisfy Opial condition [9] if for any sequence
{xn} in X, xn converges to x weakly it follows that
lim supn→∞
‖xn − x‖ < lim supn→∞
‖xn − y‖,
for all y ∈ X with y 6= x.
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces 51
Examples of Banach spaces satisfying Opial condition are Hilbert spaces and all spaces
lp(1 < p <∞). On the other hand, Lp[0, 2π] with 1 < p 6= 2 fail to satisfy Opial condition.
Definition 1.2 A multivalued mapping T : K → P (X ) is called demiclosed at y ∈ K if for any
sequence {xn} in K weakly convergent to an element x and yn ∈ T xn strongly convergent to y,
we have y ∈ T x.
The following is the multivalued version of condition (I) of Senter and Dotson [13].
Definition 1.3 A multivalued nonexpansive mapping T : K → CB(K) where K a subset of Xis said to satisfy condition (I) if there exists a nondecreasing function f : [0,∞) → [0,∞) with
f(0) = 0, f(t) > 0 for all t ∈ (0,∞) such that d(x, T x) ≥ f(d(x, F (T ))) for all x ∈ K, where
F (T ) 6= ∅ is the fixed point set of the multivalued mapping T .
We need the following Lemmas to prove our main results.
Lemma 1.4 ([18]) Let {pn}, {qn}, {rn} be three sequences of nonnegative real numbers satis-
fying the following conditions:
pn+1 ≤ (1 + qn)pn + rn, n ≥ 0,∞∑
n=0
qn <∞,∞∑
n=0
rn <∞.
Then,
(1) limn→∞
pn exists;
(2) In addition, if lim infn→∞
pn = 0, then limn→∞
pn = 0.
Lemma 1.5 ([12]) Let E be a uniformly convex Banach space and 0 < α ≤ tn ≤ β < 1 for
all n ∈ N. Suppose further that {xn} and {yn} are sequences of E such that lim supn→∞
‖xn‖ ≤ a,
lim supn→∞
‖yn‖ ≤ a and limn→∞
‖tnxn+(1−tn)yn‖ = a hold for some a ≥ 0. Then limn→∞
‖xn−yn‖= 0.
Lemma 1.6 ([16]) Let T : K → P (K) be a multivalued mapping and PT (x) = {y ∈ T x :
‖x− y‖ = d(x, T x)}. Then the following are equivalent:
(1) x ∈ F (T );
(2) PT (x) = {x};(3) x ∈ F (PT ).
Moreover, F (T ) = F (PT ).
§2. Main Results
In this section we prove some strong and a weak convergence theorems using iteration scheme
(1.6). First, we need the following lemmas to prove main results.
Lemma 2.1 Let X be a real Banach space and K be a nonempty closed and convex subset of
52 G.S.Saluja
X . Let T : K → P (K) be a multivalued mapping such that F (T ) 6= ∅ and PT is a nonexpansive
mapping. Let {xn} be the sequence defined by (1.6), where {αn}, {βn} and {γn} are real
sequences in (0, 1). Then limn→∞
‖xn − p‖ exists for all p ∈ F (T ).
Proof Let p ∈ F (T ). Then p ∈ PT (p) = {p} by Lemma 1.6. It follows from (1.6) that
‖zn − p‖ ≤ (1 − γn)‖xn − p‖ + γn‖un − p‖≤ (1 − γn)‖xn − p‖ + γnH(PT (xn), PT (p))
≤ (1 − γn)‖xn − p‖ + γn‖xn − p‖= ‖xn − p‖. (2.1)
Again using (1.6) and (2.1), we obtain
‖yn − p‖ ≤ (1 − βn)‖un − p‖ + βn‖wn − p‖≤ (1 − βn)H(PT (xn), PT (p)) + βnH(PT (zn), PT (p))
≤ (1 − βn)‖xn − p‖ + βn‖zn − p‖≤ (1 − βn)‖xn − p‖ + βn‖xn − p‖= ‖xn − p‖. (2.2)
Now using (1.6), (2.1) and (2.2), we obtain
‖xn+1 − p‖ ≤ (1 − αn)‖vn − p‖ + αn‖wn − p‖≤ (1 − αn)H(PT (yn), PT (p)) + αnH(PT (zn), PT (p))
≤ (1 − αn)‖yn − p‖ + αn‖zn − p‖≤ (1 − αn)‖xn − p‖ + αn‖xn − p‖= ‖xn − p‖. (2.3)
It follows from Lemma 1.4 that limn→∞
‖xn − p‖ exists for each p ∈ F (T ). This completes the
proof. 2Lemma 2.2 Let X be a uniformly convex Banach space and K be a nonempty closed and
convex subset of X . Let T : K → P (K) be a multivalued mapping such that F (T ) 6= ∅ and
PT is a nonexpansive mapping. Let {xn} be the sequence defined by (1.6), where {αn}, {βn}and {γn} are real sequences in (0, 1). Then lim
n→∞d(xn, T xn) = 0, lim
n→∞d(xn, T yn) = 0 and
limn→∞
d(xn, T zn) = 0.
Proof From Lemma 2.1, limn→∞
‖xn − p‖ exists for each p ∈ F (T ). We suppose that
limn→∞
‖xn − p‖ = l for some l ≥ 0.
Since lim supn→∞
‖un − p‖ ≤ lim supn→∞
H(PT (xn), PT (p)) ≤ lim supn→∞
‖xn − p‖ = l, so
lim supn→∞
‖un − p‖ ≤ l. (2.4)
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces 53
Again, since lim supn→∞
‖vn − p‖ ≤ lim supn→∞
H(PT (yn), PT (p)) ≤ lim supn→∞
‖yn − p‖ ≤ lim supn→∞
‖xn −p‖ = l, so
lim supn→∞
‖vn − p‖ ≤ l. (2.5)
Similarly,
lim supn→∞
‖wn − p‖ ≤ l. (2.6)
Applying Lemma 1.5, we get
limn→∞
‖un − vn‖ = 0 (2.7)
limn→∞
‖wn − un‖ = 0 (2.8)
and
limn→∞
‖vn − wn‖ = 0. (2.9)
Taking lim sup on both sides of (2.1) and (2.2), we obtain
lim supn→∞
‖zn − p‖ ≤ l (2.10)
and
lim supn→∞
‖yn − p‖ ≤ l. (2.11)
Also,
‖xn+1 − p‖ = ‖(1 − αn)vn + αnwn − p‖= ‖(vn − p) + αn(wn − vn)‖≤ ‖vn − p‖ + αn‖wn − vn‖≤ ‖vn − p‖ + ‖wn − vn‖
implies that
l ≤ lim infn→∞
‖vn − p‖. (2.12)
Combining (2.5) and (2.12), we obtain
limn→∞
‖vn − p‖ = l. (2.13)
Thus,
‖vn − p‖ ≤ ‖vn − wn‖ + ‖wn − p‖≤ ‖vn − wn‖ +H(PT (zn), PT (p))
≤ ‖vn − wn‖ + ‖zn − p‖
gives
l ≤ lim infn→∞
‖zn − p‖ (2.14)
54 G.S.Saluja
and by virtue of (2.10), we obtain
limn→∞
‖zn − p‖ = l. (2.15)
Similarly,
‖wn − p‖ ≤ ‖wn − vn‖ + ‖vn − p‖≤ ‖wn − vn‖ +H(PT (yn), PT (p))
≤ ‖wn − vn‖ + ‖yn − p‖
gives
l ≤ lim infn→∞
‖yn − p‖ (2.16)
and by virtue of (2.11), we obtain
limn→∞
‖yn − p‖ = l. (2.17)
Applying Lemma 1.5 once again, we obtain
limn→∞
‖xn − un‖ = 0. (2.18)
Notice that
‖xn − vn‖ ≤ ‖xn − un‖ + ‖un − vn‖.
Using (2.7) and (2.18), we obtain
limn→∞
‖xn − vn‖ = 0 (2.19)
and
‖xn − wn‖ ≤ ‖xn − un‖ + ‖un − wn‖.
Using (2.8) and (2.19), we obtain
limn→∞
‖xn − wn‖ = 0. (2.20)
Since d(xn, T xn) ≤ ‖xn − un‖, we have
limn→∞
d(xn, T xn) = 0. (2.21)
Again since d(xn, T yn) ≤ ‖xn − vn‖, we have
limn→∞
d(xn, T yn) = 0. (2.22)
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces 55
Similarly, since d(xn, T zn) ≤ ‖xn − wn‖, we have
limn→∞
d(xn, T zn) = 0. (2.23)
This completes the proof. 2Now we shall prove some strong convergence theorems using iteration scheme (1.6) in real
Banach spaces.
Theorem 2.3 Let X be a real Banach space and K be a nonempty closed and convex subset
of X . Let T : K → P (K) be a multivalued mapping such that F (T ) 6= ∅ and PT is a non-
expansive mapping. Let {xn} be the sequence defined by (??), where {αn}, {βn} and {γn}are real sequences in (0, 1). Then {xn} converges strongly to a fixed point of T if and only if
lim infn→∞
d(xn, F (T )) = 0.
Proof The necessity is obvious. Conversely, suppose that lim infn→∞
d(xn, F (T )) = 0. As
proved in Lemma 1.6, we have
‖xn+1 − p‖ ≤ ‖xn − p‖
which gives
d(xn+1, F (T )) ≤ d(xn, F (T )).
This implies that limn→∞
d(xn, F (T )) exists by Lemma 1.4 and so by hypothesis, lim infn→∞
d(xn, F (T )) =
0. Therefore, we must have limn→∞
d(xn, F (T )) = 0.
Next, we have to show that {xn} is a Cauchy sequence in K. Let ε > 0 be arbitrary chosen.
Since limn→∞
d(xn, F (T )) = 0, there exists a constant n1 such that for all n ≥ n1 we have
d(xn, F (T )) <ε
4.
In particular, inf{‖xn1 − p‖ : p ∈ F (T )} < ε4 . There must exists a q ∈ F (T ) such that
‖xn1 − q‖ < ε
2.
Now for m,n ≥ n1, we have
‖xn+m − xn‖ ≤ ‖xn+m − q‖ + ‖xn − q‖≤ 2‖xn1 − q‖< 2
(ε2
)= ε.
Hence {xn} is a Cauchy sequence in a closed subset K of a Banach space X , and so it must
56 G.S.Saluja
converge in K. Let limn→∞
xn = q1. Now
d(q1, PT (q1)) ≤ ‖xn − q1‖ + d(xn, PT (xn)) +H(PT (xn), PT (q1))
≤ 2‖xn − q1‖ + ‖xn − un‖→ 0 as n→ ∞
which gives that d(q1, T q1) = 0. But PT is a nonexpansive mapping and so F (T ) is closed.
Therefore, q1 ∈ F (PT ) = F (T ). This shows that {xn} converges strongly to a point of F (T ).
This completes the proof. 2Theorem 2.4 Let X be a real Banach space and K be a nonempty compact convex subset of
X . Let T : K → P (K) be a multivalued mapping such that F (T ) 6= ∅ and PT is a nonexpansive
mapping. Let {xn} be the sequence defined by (??), where {αn}, {βn} and {γn} are real
sequences in (0, 1). Then {xn}, {yn} and {zn} converges strongly to a fixed point of T .
Proof By Lemma 2.2, we have limn→∞
d(xn, T xn) = 0. Since by hypothesis K be a nonempty
compact convex subset of X , so there exists a subsequence {xnk} of {xn} such that lim
k→∞‖xnk
−u‖ = 0 for some u ∈ K. Thus
d(u, PT (u)) ≤ ‖xnk− u‖ + d(xnk
, PT (xnk)) +H(PT (xnk
), PT (u))
≤ 2‖xnk− u‖ + ‖xnk
− unk‖
→ 0 as k → ∞.
This shows that u is a fixed of T . From Lemma 2.1, we get that limn→∞
‖xn − u‖ = 0. Again
from Lemma 2.2, we get that
‖yn − xn‖ = ‖(1 − βn)un + βnwn − xn‖≤ ‖un − xn‖ + βn‖wn − un‖≤ ‖un − xn‖ + ‖wn − un‖→ 0 as n→ ∞,
and
‖zn − xn‖ = ‖(1 − γn)xn + γnun − xn‖≤ γn‖un − xn‖≤ ‖un − xn‖→ 0 as n→ ∞.
It follows that limn→∞
‖yn − u‖ = 0 and limn→∞
‖zn − u‖ = 0. Thus the conclusion follows. This
completes the proof. 2Now, we apply Theorem 2.3 to obtain another strong convergence theorem in uniformly
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces 57
convex Banach spaces satisfies condition (I) of Senter and Dotson [13].
Theorem 2.5 Let X be a uniformly convex Banach space and K be a nonempty closed and
convex subset of X . Let T : K → P (K) be a multivalued mapping satisfying condition (I) such
that F (T ) 6= ∅ and PT is a nonexpansive mapping. Let {xn} be the sequence defined by (1.6),
where {αn}, {βn} and {γn} are real sequences in (0, 1). Then {xn} converges strongly to a fixed
point of T .
Proof By Lemma 2.1, limn→∞
‖xn − p‖ exists for all p ∈ F (T ) and so the sequence {xn} is
bounded . Let limn→∞
‖xn − p‖ = r for some r > 0.
Now from Lemma 2.1, we know that
‖xn+1 − p‖ ≤ ‖xn − p‖
which implies that
infp∈F (T )
‖xn+1 − p‖ ≤ infp∈F (T )
‖xn − p‖,
and also d(xn+1, F (T )) ≤ d(xn, F (T )). And so, limn→∞
d(xn, F (T )) exists. By using condition
(I) and Lemma 2.2, we have
limn→∞
f(d(xn, F (T ))) ≤ limn→∞
d(xn, T xn) = 0.
That is,
limn→∞
f(d(xn, F (T ))) = 0.
Since f is a nondecreasing function and f(0) = 0, it follows that limn→∞
d(xn, F (T )) = 0. The
conclusion follows from Theorem 2.3. This completes the proof. 2Now, we prove the weak convergence theorem of the sequence {xn} defined by (1.6).
Theorem 2.6 Let X be a uniformly convex Banach space satisfying Opial’s condition and K be
a nonempty closed and convex subset of X . Let T : K → P (K) be a multivalued mapping such
that F (T ) 6= ∅ and PT is a nonexpansive mapping. Let {xn} be the sequence defined by (??),
where {αn}, {βn} and {γn} are real sequences in (0, 1). Let I − PT be demiclosed with respect
to zero. Then {xn} converges weakly to a fixed point of T .
Proof Let z ∈ F (T ). From Lemma 2.1, we know that limn→∞
‖xn − z‖ exists. Now we
prove that {xn} has a unique weak subsequential limit in F (T ). To prove this, let p1 and p2 be
weak limits of the subsequences {xni} and {xnj
} of {xn}, respectively. By (2.18), there exists
un ∈ T xn such that limn→∞
‖xn − un‖ = 0. Since I − PT is demiclosed with respect to zero,
therefore we obtain p1 ∈ F (T ). In the same way, we can prove that p2 ∈ F (T ).
Next, we prove uniqueness. For this, suppose that p1 6= p2. Then by Opial’s condition, we
58 G.S.Saluja
have
limn→∞
‖xn − p1‖ = limni→∞
‖xni− p1‖
< limni→∞
‖xni− p2‖
= limn→∞
‖xn − p2‖= lim
nj→∞‖xnj
− p2‖
< limnj→∞
‖xnj− p1‖
= limn→∞
‖xn − p1‖,
which is a contradiction. Hence {xn} converges weakly to a fixed point of T . This completes
the proof. 2Remark 2.7 Our results extend, generalize and improve several corresponding results from
the existing literature and iterative schemes discussed by Panyanak [10], Sastry and Babu [11],
Song and Wang [14] and Song and Cho [16] in the sense of faster iterative scheme.
Suzuki [17] introduced a condition on mappings called condition (C) which is weaker than
nonexpansiveness.
Recently, Abkar and Eslamian [2] introduced the definition of condition (C) for multi-
valued mapping. The definition is as follows.
Definition 2.8([2]) A multivalued mapping T : X → CB(X ) is said to satisfy condition (C)
provided that1
2d(x, T x) ≤ ‖x− y‖ ⇒ H(T x, T y) ≤ ‖x− y‖, x, y ∈ X .
The following result can be found in [2].
Lemma 2.9([2]) Let T : X → CB(X ) be a multi-valued mapping. If T is nonexpansive, then
T satisfies the condition (C).
We mention that there exist single-valued and multi-valued mappings satisfying the con-
dition (C) which are not nonexpansive.
Example 2.10([17]) Define a mapping T on [0.3] by
T (x) =
0, if x 6= 3,
1, if x = 3.
Then T is a single-valued mapping satisfying condition (C), but T is not nonexpansive.
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces 59
Example 2.11([2]) Define a mapping T : [0, 5] → [0, 5] by
T (x) =
[0, x5 ], if x 6= 5,
{1}, if x = 5.
Then it is easy to show that T is a multi-valued mapping satisfying condition (C), but T is not
nonexpansive.
Now, we obtain some strong convergence results using iteration scheme (1.6) and condition
(C).
Theorem 2.12 Let X be a real Banach space and K be a nonempty closed and convex subset of
X . Let T : K → P (K) be a multivalued mapping such that F (T ) 6= ∅ and PT satisfies condition
(C). Let {xn} be the sequence defined by (1.6), where {αn}, {βn} and {γn} are real sequences in
(0, 1). Then {xn} converges strongly to a fixed point of T if and only if lim infn→∞
d(xn, F (T )) = 0.
Proof The proof of Theorem 2.12 immediately follows from Lemma 2.1 and Theorem 2.3.
This completes the proof. 2Theorem 2.13 Let X be a real Banach space and K be a nonempty closed and convex subset of
X . Let T : K → P (K) be a multivalued mapping such that F (T ) 6= ∅ and PT satisfies condition
(C). Let {xn} be the sequence defined by (1.6), where {αn}, {βn} and {γn} are real sequences
in (0, 1). If the following condition is satisfied:
(c1) there exists an increasing function ψ : [0,∞) → [0,∞) with ψ(r) > 0, ∀ r > 0 such
that
d(xn, T xn) ≥ ψ(d(xn, F (T )), (2.24)
then {xn} converges strongly to a fixed point of T .
Proof As in the proof of Lemma 2.2, we know that limn→∞
d(xn, T xn) = 0. Hence from
(2.21) we obtain limn→∞
d(xn, F (T )) = 0. The conclusion of Theorem 2.13 can be obtained from
Theorem 2.12 immediately. This completes the proof. 2§3. Concluding Remarks
In this paper, we study a new three-step iteration scheme for multivalued nonexpansive map-
pings in Banach spaces and establish some strong convergence theorems and a weak convergence
theorem under some appropriate conditions applying on the space. Our results extend and gen-
eralize several results from the current existing literature.
References
[1] M. Abbas and T. Nazir, A new faster iteration process applied to constrained minimization
and feasibility problems, Mat. Vesnik, 66(2) (2014), 223-234.
[2] A. Abkar and M. Eslamian, Fixed point theorems for Suzuki generalized nonexpansive
60 G.S.Saluja
multivalued mappings in Banach spaces, Fixed Point Theory Appl., (2010) Article ID
457935, 10 pages.
[3] R. P. Agarwal, D. O’Regan and D. R. Sahu, Iterative construction of fixed points of nearly
asymptotically nonexpansive mappings, Nonlinear Convex Anal., 8(1) (2007), 61-79.
[4] N. A. Assad and W. A. Kirk, Fixed point theorems for set-valued mappings ocontractive
type, Pacific J. Math., 43 (1972), 553-562.
[5] L. Gorniewicz, Topologfixed Point Theory of Multivalued Mappings, Kluwer Academic Pub.
Dordrecht, Netherlands, (1999).
[6] W. A. Kirk, Transfinite methods in metric fixed point theory, Abstract and Applied Anal-
ysis, 5 (2003), 311-324.
[7] T. C. Lim, A fixed point theorem for multivalued nonexpansive mappings in a uniformly
convex Banach space, Bull. Amer. Math. Soc., 80 (1974), 1123-1126.
[8] T. C. Lim, Remarks on some fixed point theorems, Proc. Amer. Math. Soc., 60 (1976),
179-182.
[9] Z. Opial, Weak convergence of the sequence of successive approximations for nonexpansive
mappings, Bull. Amer. Math. Soc., 73 (1967), 591-597.
[10] B. Panyanak, Mann and Ishikawa iterative processes for multivalued mappings in Banach
spaces, Comp. Math. Appl., 54 (2007), 872-877.
[11] K. P. R. Sastry, and G. V. R. Babu, Convergence of Ishikawa iterates for a multivalued
mapping with a fixed point, Czechoslovak Math. J., 55 (2005), 817-826.
[12] J. Schu, Iterative construction of fixed points of asymptotically nonexpansive mappings,
J. Math. Anal. Appl., 158 (1991), 407-413.
[13] H. F. Senter and W. G. Dotson, Approximating fixed points of nonexpansive mappings,
Proc. Amer. Math. Soc., 44 (1974), 375-380.
[14] Y. S. Song and H. J. Wang, Erratum Mann and Ishikawa iterative processes for multivalued
mappings in Banach spaces, Comput. Math. Appl., 55 (2008), 2999-3002.
[15] Y. S. Song and H. J. Wang, Convergence of iterative algorithms for multivalued mappings
in Banach spaces, Nonlinear Anal., 70(4) (2009), 1547-1556.
[16] Y. S. Song and Y. J. Cho, Some notes on Ishikawa iteration for multivaled mappings, Bull.
Korean Math. Soc., 48(3) (2011), 575-584.
[17] T. Suzuki, Fixed point theorems and convergence theorems for some generalized nonex-
pansive mappings, J. Math. Anal. Appl., 340(2) (2008), 1088-1095.
[18] K. K. Tan, H. K. Xu, Approximating fixed points of nonexpansive mappings by the Ishikawa
iteration process, J. Math. Anal. Appl., 178 (1993), 301-308.
[19] H. K. Xu, Multivalued nonexpansive mappings in Banach spaces, Nonlinear Anal., 43
(2001), 693-706.
International J.Math. Combin. Vol.4(2019), 61-69
Algebraic Properties of the Path Complexes of Cycles
Seyed Mohammad Ajdani
(Department of Mathematics, Zanjan Branch, Islamic Azad University, Zanjan, Iran)
E-mail: [email protected]
Abstract: Let G be a simple graph and ∆t (G) be a simplicial complex whose facets
correspond to the paths of length t (t ≥ 2) in G. It is shown that ∆t (Cn) is matroid, vertex
decomposable, shellable and Cohen-Macaulay if and only if n = t or n = t + 1, where Cn is
an n-cycle. As a consequence we show that if n = t or t + 1 then ∆t (Cn) is partitionable
and Stanley’s conjecture holds for K[∆t (Cn)].
Key Words: Vertex decomposable, simplicial complex, matroid, path.
AMS(2010): 13F20, 05E40, 13F55.
§1. Introduction
Let R = K[x1, · · · , xn], where K is a field. Fix an integer n ≥ t ≥ 2 and let G be a directed
graph. A sequence xi1 , · · · , xitof distinct vertices is called a path of length t if there are t− 1
distinct directed edges e1, · · · , et−1 where ej is a directed edge from xijto xij+1 . Then the path
ideal of G of length t is the monomial ideal It(G) = (xi1 · · ·xit: xi1 , · · · , xit
is a path of length
t in G in the polynomial ring R = K[x1, · · · , xn]. The distance d(x, y) of two vertices x and y
of a graph G is the length of the shortest path from x to y. The path complex ∆t(G) is defined
by
∆t(G) = 〈{xi1 , · · · , xit} : xi1 , · · · , xit
is a path of length t in G 〉.
Path ideals of graphs were first introduced by Conca and De Negri [3] in the context of
monomial ideals of linear type. Recently the path ideal of cycles has been extensively studied
by several mathematicians. In [9], it is shown that I2(Cn) is sequentially Cohen-Macaulay, if
and only if, n = 3 or n = 5. Generalizing this result, in [13], it is proved that It(Cn), (t > 2),
is sequentially Cohen-Macaulay, if and only if n = t or n = t+ 1 or n = 2t+ 1. Also, the Betti
numbers of the ideal It(Cn) and It(Ln) is computed explicitly in [1]. In particular, it has been
shown that
Theorem 1.1(Corollary 5.15, [1]) Let n, t, p and d be integers such that n ≥ t ≥ 2, n =
(t+ 1)p+ d, where p ≥ 0 and 0 ≤ d < (t+ 1). Then,
1Received June 17, 2019, Accepted November 26, 2019.
62 Seyed Mohammad Ajdani
(i) The projective dimension of the path ideal of a graph cycle Cn or line Ln is given by
pd (It(Cn)) =
2p, d 6= 0
2p− 1, d = 0pd (It(Ln)) =
2p− 1, d 6= t,
2p, d = t.
(ii) The regularity of the path ideal of a graph cycle Cn or line Ln is given by
reg (It(Cn)) = (t− 1)p+ d+ 1
reg (It(Ln)) =
p(t− 1) + 1, d < t,
p(t− 1) + t, d = t.
In [8] it has been shown that, ∆t(G) is a simplicial tree if G is a rooted tree and t ≥ 2.
One of interesting problems in combinatorial commutative algebra is the Stanley’s conjectures.
The Stanley’s conjectures are studied by many researchers. Let R be a Nn-graded ring and M
a Zn- graded R- module. Then, Stanley [10] conjectured that
depth (M) ≤ sdepth (M)
He also conjectured in [11] that each Cohen-Macaulay simplicial complex is partitionable. Her-
zog, Soleyman Jahan and Yassemi in [7] showed that the conjecture about partitionability is
a special case of the Stanley’s first conjecture. In this work, we study algebraic properties of
∆t(Cn). In Section 1, we recall some definitions and results which will be needed later. In
Section 3, we show that the following conditions are equivalent for all t > 2:
(i) ∆t(Cn) is matroid;
(ii) ∆t(Cn) is vertex decomposable;
(iii) ∆t(C n) is shellable;
(iv) ∆t(Cn) is Cohen-Macaulay;
(v) n = t or t+ 1.
(See Theorem 3.6).
In Section 4 as an application of our results we show that if n = t or t+ 1 then ∆t (Cn) is
partitionable and Stanley’s conjecture holds for K[∆t (Cn)].
§2. Preliminaries
In this section we recall some definitions and results which will be needed later.
Definition 2.1 A simplicial complex ∆ over a set of vertices V = {x1, · · · , xn}, is a collection
of subsets of V , with the property that:
(a) {xi} ∈ ∆ for all i;
(b) If F ∈ ∆, then all subsets of F are also in ∆ (including the empty set).
Algebraic Properties of the Path Complexes of Cycles 63
An element of ∆ is called a face of ∆ and complement of a face F is V \F and it is denoted
by F c. Also, the complement of the simplicial complex ∆ = 〈F1, · · · , Fr〉 is ∆c = 〈F c1 , · · · , F c
r 〉.The dimension of a face F of ∆, dimF , is |F | − 1 where, |F | is the number of elements of F
and dim ∅ = −1. The faces of dimensions 0 and 1 are called vertices and edges, respectively.
A non-face of ∆ is a subset F of V with F /∈ ∆. we denote by N (∆), the set of all minimal
non-faces of ∆. The maximal faces of ∆ under inclusion are called facets of ∆. The dimension
of the simplicial complex ∆, dim∆, is the maximum of dimensions of its facets. If all facets of
∆ have the same dimension, then ∆ is called pure.
Let F(∆) = {F1, · · · , Fq} be the facet set of ∆. It is clear that F(∆) determines ∆
completely and we write ∆ = 〈F1, · · · , Fq〉. A simplicial complex with only one facet is called
a simplex. A simplicial complex Γ is called a subcomplex of ∆, if F(Γ) ⊂ F(∆).
For v ∈ V , the subcomplex of ∆ obtained by removing all faces F ∈ ∆ with v ∈ F is
denoted by ∆ \ v. That is,
∆ \ v = 〈F ∈ ∆ : v /∈ F 〉.
The link of a face F ∈ ∆, denoted by link∆(F ), is a simplicial complex on V with the
faces, G ∈ ∆ such that, G ∩ F = ∅ and G ∪ F ∈ ∆. The link of a vertex v ∈ V is simply
denoted by link∆(v).
link∆(v) ={F ∈ ∆ : v /∈ F, F ∪ {v} ∈ ∆
}.
Let ∆ be a simplicial complex over n vertices {x1, · · · , xn}. For F ⊂ {x1, · · · , xn}, we set:
xF =∏
xi∈F
xi.
We define the facet ideal of ∆, denoted by I(∆), to be the ideal of S generated by {xF : F ∈F(∆)}. The non-face ideal or the Stanley-Reisner ideal of ∆, denoted by I∆, is the ideal of
S generated by square-free monomials {xF : F ∈ N (∆)}. Also we call K[∆] := S/I∆ the
Stanley-Reisner ring of ∆.
Definition 2.2 A simplicial complex ∆ on {x1, · · · , xn} is said to be a matroid if, for any two
facets F and G of ∆ and any xi ∈ F , there exists a xj ∈ G such that (F \ {xi}) ∪ {xj} is a
facet of ∆.
Definition 2.3 A simplicial complex ∆ is recursively defined to be vertex decomposable, if it is
either a simplex, or else has some vertex v so that
(a) Both ∆ \ v and link∆(v) are vertex decomposable, and
(b) No face of link∆(v) is a facet of ∆ \ v.A vertex v which satisfies in condition (b) is called a shedding vertex.
Definition 2.4 A simplicial complex ∆ is shellable, if the facets of ∆ can be ordered F1, · · · , Fs
such that, for all 1 ≤ i < j ≤ s, there exists some v ∈ Fj \ Fi and some l ∈ {1, · · · , j − 1} with
Fj \ Fl = {v}.
64 Seyed Mohammad Ajdani
A simplicial complex ∆ is called disconnected, if the vertex set V of ∆ is a disjoint union
V = V1 ∪ V2 such that no face of ∆ has vertices in both V1 and V2. Otherwise ∆ is connected.
It is well-known that
matroid =⇒ vertex decomposable =⇒ shellable =⇒ Cohen-Macaulay
Definition 2.5 For a given simplicial complex ∆ on V , we define ∆∨, the Alexander dual of
∆, by
∆∨ = {V \ F : F /∈ ∆}.
It is known that for the complex ∆ one has I∆∨ = I(∆c). Let I 6= 0 be a homogeneous
ideal of S and N be the set of non-negative integers. For every i ∈ N ∪ {0}, one defines:
tSi (I) = max{j : βSi,j(I) 6= 0}
where βSi,j(I) is the i, j-th graded Betti number of I as an S-module. The Castelnuovo-Mumford
regularity of I is given by:
reg(I) = sup{tSi (I) − i : i ∈ Z}.
We say that the ideal I has a d-linear resolution, if I is generated by homogeneous polynomials
of degree d and βSi,j(I) = 0, for all j 6= i + d and i ≥ 0. For an ideal which has a d-linear
resolution, the Castelnuovo-Mumford regularity would be d. If I is a graded ideal of S, we
write (Id) for the ideal generated by all homogeneous polynomials of degree d belonging to I.
Definition 2.6 A graded ideal I is componentwise linear if (Id) has a linear resolution for all
d.
Also, we write I[d] for the ideal generated by the squarefree monomials of degree d belonging
to I.
Definition 2.7 A graded S-module M is called sequentially Cohen-Macaulay (over K), if there
exists a finite filtration of graded S-modules,
0 = M0 ⊂M1 ⊂ · · · ⊂Mr = M
such that each Mi/Mi−1 is Cohen-Macaulay, and the Krull dimensions of the quotients are
increasing:
dim(M1/M0) < dim(M2/M1) < · · · < dim(Mr/Mr−1).
The Alexander dual, allows us to make a bridge between (sequentially) Cohen-Macaulay
ideals and (componetwise) linear ideals.
Definition 2.8(Alexander Duality) For a square-free monomial ideal I = (M1, · · · ,Mq) ⊂ S =
Algebraic Properties of the Path Complexes of Cycles 65
K[x1, · · · , xn], the Alexander dual of I, denoted by I∨, is defined to be
I∨ = PM1 ∩ · · · ∩ PMq
where, PMiis prime ideal generated by {xj : xj |Mi}.
Theorem 2.9(Proposition 8.2.20, [6]; Theorem 3, [4]) Let I be a square-free monomial ideal
in S = K[x1, · · · , xn].
(i) The ideal I is componentwise linear ideal if and only if S/I∨ is sequentially Cohen-
Macaulay;
(ii) The ideal I has a q-linear resolution if and only if S/I∨ is Cohen-Macaulay of dimen-
sion n− q.
Remark 2.10 Two special cases, we will be considering in this paper, are when G is a cycle
Cn, or a line graph Ln on vertices {x1, · · · , xn} with edges
E (Cn) ={{x1, x2}, {x2, x3}, · · · , {xn−1, xn}, {xn, x1}
};
E (Ln) ={{x1, x2}, {x2, x3}, · · · , {xn−1, xn}
}.
§3. Vertex Decomposability Path Complexes of Cycles
As the main result of this section, it is shown that ∆t (Cn) is matroid, vertex decomposable,
shellable and Cohen-Macaualay if and only if n = t or n = t + 1. For the proof we shall need
the following lemmas and propositions.
Lemma 3.1 Let ∆t(Ln) be a simplicial complex on {x1, · · · , xn} and 2 ≤ t ≤ n. Then ∆t(Ln)
is vertex decomposable.
Proof If t = n, then ∆n(Ln) is a simplex which is vertex decomposable. Let 2 ≤ t < n
then one has
∆t(Ln) = 〈{x1, · · · , xt}, {x2, · · · , xt+1}, · · · , {xn−t+1, · · · , xn}〉.
So ∆t(Ln) \ xn = 〈{x1, · · · , xt}, {x2, · · · , xt+1}, · · · , {xn−t, · · · , xn−1}〉. Now we use induction
on the number of vertices of Ln and by induction hypothesis ∆t(Ln)\xn is vertex decomposable.
On the other hand, it is clear that link∆t(Ln){xn} = 〈{xn−t+1, · · · , xn−1}〉. Thus link∆t(Ln){xn}is a simplex which is not a facet of ∆t(Ln) \ xn. Therefore ∆t(Ln) is vertex decomposable. 2Lemma 3.2 Let ∆2(Cn) be a simplicial complex on {x1, · · · , xn}. Then ∆2(Cn) is vertex
decomposable.
Proof Since ∆2(Cn) = 〈{x1, x2}, {x2, x3}, · · · , {xn−1, xn}, {xn, x1}〉, we have
∆2(Cn) \ xn = 〈{x1, , x2}, {x2, x3}, · · · , {xn−2, xn−1}〉.
66 Seyed Mohammad Ajdani
By lemma 3.1 ∆2(Cn)\xn is vertex decomposable. Also it is trivial that link∆2(Cn){xn} =
〈{xn−1}, {x1}〉 is vertex decomposable and no face of link∆2(Cn){xn} is a facet of ∆2(Cn) \ xn.
Therefore ∆2(Cn) is vertex decomposable. 2Lemma 3.3 Let ∆t(Cn) be a simplicial complex on {x1, · · · , xn} and 3 ≤ t ≤ n − 2. Then
∆t(Cn) is not Cohen-Macaulay.
Proof It suffices to show that I∆t(Cn)∨ has not a linear resolution. Since I∆t(Cn)∨ =
I(∆t(Cn)c) then one can easily check that I∆t(Cn)∨ = In−t(Cn). By Theorem 1.1 we have
reg(I∆t(Cn)∨) = (n− t− 1)p+ d+ 1.
Since 3 ≤ t ≤ n− 2 then one has reg(I∆t(Cn)∨) 6= n− t and by Theorem 2.9 ∆t(Cn) is not
Cohen-Macaulay. 2Proposition 3.4 Let ∆t(Cn) be a simplicial complex on {x1, · · · , xn} and t ≥ 3. Then ∆t(Cn)
is vertex decomposable if and only if n = t or t+ 1.
Proof By Lemma 3.3 it suffices to show that if n = t or t + 1, then ∆t(Cn) is vertex
decomposable. If n = t, then ∆n(Cn) is a simplex which is vertex decomposable. If t = n− 1,
then we have
∆n−1(Cn) = 〈{x1, · · · , xn−1}, {x2, · · · , xn}, {x3, · · · , xn, x1}, · · · , {xn, x1, · · · , xn−2}〉.
Now we use induction on the number of vertices of Cn and show that ∆n−1(Cn) is vertex
decomposable. It is clear that ∆n−1(Cn) \ xn = 〈{x1, · · · , xn−1}〉 is a simplex which is vertex
decomposable.
On the other hand,
link∆n−1(Cn){xn} = 〈{x1, · · · , xn−2}, · · · , {xn−1, x1, · · · , xn−3}〉 = ∆n−2(Cn−1).
By induction hypothesis link∆n−1(Cn){xn} is vertex decomposable. It is easy to see that
no face of link∆n−1(Cn){xn} is a facet of ∆n−1(Cn) \ xn. Therefore ∆n−1(Cn) is vertex decom-
posable. 2Proposition 3.5 ∆2(Cn) is a matroid if and only if n = 3 or 4.
Proof If n = 3 or 4, then it is easy to see that ∆2(Cn) is a matroid. Now we prove the
converse. It suffices to show that ∆2(Cn) is not a matroid for all n ≥ 5. We consider two
facets {x1, x2} and {xn−1, xn}. Then we have ({x1, x2} \ {x1}) ∪ {xn−1} = {x2, xn−1} and
({x1, x2}\{x1})∪{xn} = {x2, xn}. Since {x2, xn−1} and {x2, xn} are not the facets of ∆2(Cn).
So ∆2(Cn) is not matroid for all n ≥ 5. 2For the simplicial complexes one has the following implication:
Matroid ⇒ vertex decomposable ⇒ shellable ⇒ Cohen-Macaulay
Algebraic Properties of the Path Complexes of Cycles 67
Note that these implications are strict, but by the following theorem, for path complexes,
the reverse implications are also valid.
Theorem 3.6 Let t ≥ 3. Then the following conditions are equivalent:
(i) ∆t(Cn) is matroid;
(ii) ∆t(Cn) is vertex decomposable;
(iii) ∆t(Cn) is shellable;
(iv) ∆t(Cn) is Cohen-Macaulay;
(v) n = t or t+ 1.
Proof (i) =⇒ (ii), (ii) =⇒ (iii) and (iii) =⇒ (iv) is well-known.
(iv) =⇒ (v) follows from Lemma 3.3 and Proposition 3.4.
(v) =⇒ (i): If n = t, then ∆t(Cn) is a simplex which is a matroid. If n = t+ 1, then
∆t(Cn) = 〈{x1, · · · , xt}, {x2, · · · , xt+1}, {x3, · · · , xt+1, x1}, · · · , {xt+1, x1, · · · , xt−1}〉.
For any two facets F and G of ∆t(Cn) one has | F ∩G |= t− 1. We claim that for any two
facets F and G of ∆t(Cn) and any xi ∈ F , there exists a xj ∈ G such that (F \{xi})∪{xj} is a
facet of ∆t(Cn). We have to consider two cases. If xi ∈ F and xi /∈ G, then we choose xj ∈ G
such that xj /∈ F . Thus (F \ {xi}) ∪ {xj} = G which is a facet of ∆t(Cn).
For other case, if xi ∈ F and xi ∈ G, then we choose xj ∈ G such that xj is the same xi.
Therefore (F \ {xi}) ∪ {xi} = F is a facet of ∆t(Cn) which completes the proof. 2§4. Stanley Decompositions
Let R be any standard graded K- algebra over an infinite field K, i.e, R is a finitely gener-
ated graded algebra R =⊕
i≥0 Ri such that R0 = K and R is generated by R1. There are
several characterizations of the depth of such an algebra. We use the one that depth (R) is
the maximal length of a regular R- sequence consisting of linear forms. Let xF = ⊓i∈Fxi be
a squarefree monomial for some F ⊆ [n] and Z ⊆ {x1, · · · , xn}. The K- subspace xFK[Z]
of S = K[x1, · · · , xn] is the subspace generated by monomials xFu, where u is a monomial in
the polynomial ring K[Z]. It is called a square free Stanley space if {xi : i ∈ F} ⊆ Z. The
dimension of this Stanley space is |Z|. Let ∆ be a simplicial complex on {x1, · · · , xn}. A square
free Stanley decomposition D of K[∆] is a finite direct sum⊕
i uiK[Zi] of squarefree Stanley
spaces which is isomorphic as a Zn- graded K- vector space to K[∆], i.e.
K[∆] ∼=⊕
i
uiK[Zi].
We denote by sdepth (D) the minimal dimension of a Stanley space in D and define sdepth (K[∆])
= max{sdepth (D)}, where D is a Stanley decomposition of K[∆]. Stanley conjectured in [10]
68 Seyed Mohammad Ajdani
the upper bound for the depth of K[∆] holding with
depth (K[∆]) ≤ sdepth (K[∆]).
Also we recall another conjecture of Stanley. Let ∆ be again a simplicial complex on
{x1, · · · , xn} with facets G1, · · · , Gt. The complex ∆ is called partitionable if there exists a
partition ∆ =⋃t
i=1[Fi, Gi] where Fi ⊆ Gi are suitable faces of ∆. Here the interval [Fi, Gi] is
the set of faces {H ∈ ∆ : Fi ⊆ H ⊆ Gi}. In [11] and [12] respectively Stanley conjectured each
Cohen-Macaulay simplicial complex is partitionable. This conjecture is a special case of the
previous conjecture. Indeed, Herzog, Soleyman Jahan and Yassemi [7] proved that for Cohen-
Macaulay simplicial complex ∆ on {x1, . . . , xn} we have that depth (K[∆]) ≤ sdepth (K[∆]) if
and only if ∆ is partitionable. Since each vertex decomposable simplicial complex is shellable
and each shellable complex is partitionable. Then as a consequence of our results, we obtain
Corollary 3.1 If n = t or t + 1 then ∆t (Cn) is partitionable and Stanley’s conjecture holds
for K[∆t (Cn)].
Acknowledgment. The author is deeply grateful to the referee for careful reading of the
manuscript and helpful suggestions.
References
[1] A. Aliloee and S. Faridi, Betti numbers of the path ideals of cycles and lines, preprint
(2011).
[2] A. Bj’orner, Topological methods, Handbook of combinatorics, Vol.1, 2, 1819-1872, Elsevier,
Amsterdam, (1995).
[3] A. Conca and E. De Negri, M-sequences, graph ideals and ladder ideals of linear type, J.
Algebra, 211 (1999), no. 2, 599-624.
[4] J. A. Eagon and V. Reiner, Resolutions of Stanley-Reisner rings and Alexander duality, J.
Pure and Applied Algebra, 130 (1998), 265-275.
[5] J. Herzog and T. Hibi, Monomial Ideals, in: GTM 260, Springer, London, (2010).
[6] J. Herzog and T. Hibi, Componentwise linear ideals, Nagoya Math. J., 153 (1999), 141-153.
[7] J. Herzog, A. Soleyman Jahan and S. Yassemi, Stanley decompositions and partitionable
simplicial complexes, J. Alger. Comb., 27(2008), 113-125
[8] J. He and A. Van Tuyl, Algebraic properties of the path ideal of a tree, Comm. Algebra.,
38 (2010), 1725-1742.
[9] C. A. Francisco and A. Van Tuyl, Sequentially Cohen-Macaulay edge ideals, Proc. Amer.
Math. Soc., (2007). No.8, 2327-2337.
[10] R. P. Stanley, Linear Diophantine equations and local cohomology, Invent. Math., 68(1982),
No 2, 175-193.
[11] R. P. Stanley, Combinatorics and Commutative Algebra(Second edition), Progress in Math-
ematics, 41, Birkhauser Boston 1996.
Algebraic Properties of the Path Complexes of Cycles 69
[12] R. p. Stanley, Positivity problems and conjectures in algebraic combinatorics, In Mathe-
matics: Frontiers and Perspectives, 295-319, Amer. Math. Soc., Providence, RI 2000.
[13] S. Saeedi Madani, D. Kiani and N. Terai, Sequentially Cohen-Macaulay path ideals of
cycles, Bull. Math. Soc. Sci. Math. Roumanie Tome, 54 (102), No. 4, (2011), 353-363.
[14] E. Sampathkumar, P.S.K Reddy, M.S Subramanya, The line n- sigraph of a Symmetric n-
sigraph, Southeast Asian Bull. Matt., 34 (5)(2010), 953-958.
[15] R. Woodroofe, Vertex decomposable graphs and obstructions to shellability, Proc. Amer.
Math. Soc., 137 (2009), 3235-3246.
International J.Math. Combin. Vol.4(2019), 70-79
Some Classes of Analytic Functions with q-Calculus
Hamid Shamsan1, Fuad S. M. Alsarari2 and S. Latha1
1. Department of Mathematics, Yuvaraja’s College, University of Mysore, Mysuru-570 005, India
2. Department of Mathematics and Statistics, College of Sciences, Yanbu, Taibah University, Saudi Arabia
E-mail: [email protected], [email protected], [email protected]
Abstract: We study the estimates for the Second q-Hankel determinant of analytic func-
tions in a class which unifies a number of classes studied previously by Darus, Ramreddy,
Ravichandran, Yang and others. Our class includes q-convex and q-starlike functions. Also
we study the estimate for q-Toeplitz determinants whose elements are the coefficients an for
f in close-to-q-convex functions.
Key Words: Second Hankel determinant, subordination, q-starlike and q-convex functions,
close-to-q-convex, Toeplitz matrices.
AMS(2010): 30C45. 30C50.
§1. Introduction
The Hankel determinants Hq(n) of Taylor’s coefficients of function f ∈ A where A denotes the
class of functions of the form
f(z) = z +
∞∑
n=2
anzn, (1.1)
which are analytic in the open unit disk U = {z : z ∈ C and |z| < 1}. is defined by
Hq(n) :=
∣∣∣∣∣∣∣∣∣∣∣∣
an an+1 · · · an+q−1
an+1 an+2 · · · an+q
...... · · ·
...
an+q−1 an+q · · · an+2q−2
∣∣∣∣∣∣∣∣∣∣∣∣
where (a1 = 1, n = q ∈ N). Hankel matrices (and determinants) play an important role in
several branches of mathematics and have many applications [11]. H2(1) is the classical Fekete-
Szego functional. Fekete-Szego in [4] found the maximum value of H2(1). Pommerenke in [16]
proved that the Hankel determinant of univalent functions satisfy
|Hq(n)| < Kn−( 12 +β)q+ 3
2 (n = 1, 2, · · · , q = 2, 3, · · · ),
1Received July 31, 2019, Accepted December 1, 2019.
Some Classes of Analytic Functions with q-Calculus 71
where β > 14000 and K depends only on q.
Hayman [8] showed that
|H2(n)| = |anan+2 − a2n+1| < An
12 , n = 2, 3, · · · ,
where A is an absolute constant for a really mean univalent functions. Hankel determinants
are useful in showing that a function of bounded characteristic in U , i.e, a function which
is ratio of tow bounded analytic functions with its Laurent series around the origin having
integral coefficients, is rational. In recent years, several authors investigated bounds for the
Hankel determinant belonging tow various subclasses of univalent and multivalent functions
in a class which unifies a number of classes studied earlier by Deepak Bansal, K. I. Noor, T.
Yavuz, Sarika Verma, Shigeyoshi Owa and others. Closely related to Hankel determinants are
the Toepliz determinants. A Toeplitz matrix Tq(n) defined by
Tq(n) :=
∣∣∣∣∣∣∣∣∣∣∣∣
an an+1 · · · an+q−1
an+1 an · · ·...
... · · ·...
an+q−1 an+q · · · an
∣∣∣∣∣∣∣∣∣∣∣∣
A Toeplitz matrix can be thought of as an upside-down Hankel matrix, in that Hankel ma-
trices have constant entries along the reverse diagonal, whereas Toeplitz matrices have constant
entries along the diagonal. A good summary of the applications of Toeplitz matrices to a wide
range of areas of pure and applied mathematics can also be found in [11].
We aim to define q-starlike, q-convex functions and Ma-Minda starlike and convex func-
tions. We use the concept of principle of subordination and q-calculus to define our classes.
Recently in the second half of the twentieth century q-calculus aroused interest due to lot of
applications in the various mathematical fields such as combinatorics, number theory, quantum
theory and the theory of relativity. The q-derivative of a function is defined in the following.
Definition 1.1([9]) The q-derivative of f is given by
∂qf(z) =
f(z)−f(qz)z(1−q) , z 6= 0,
f ′(0), z = 0., where 0 < q < 1. (1.2)
Equivalently, (1.2) may be written as
∂qf(z) = 1 +
∞∑
n=2
[n]qanzn−1, z 6= 0
where
[n]q =
1−qn
1−q, q 6= 1
n, q= 1.
72 Hamid Shamsan, Fuad S. M. Alsarari and S. Latha
Note that as q → 1−, [n]q → n.
Definition 1.2([7]) Let f be analytic in U and be given by (1.1). Then a function f is starlike
if and only if, ℜ{
zf ′(z)f(z)
}> 0. We denote the class of starlike functions by S∗.
The class of functions with positive real part plays a significant role in complex function
theory. Using principle of subordination we define the functions with positive real part.
Definition 1.3([17]) Let f and g be analytic in U , then f is said to be subordinate to the
function g, written f(z) ≺ g(z), if there exists an analytic function ω : U → U satisfying
ω(0) = 0 and |ω(z)| < 1 such that f(z) = g(ω(z)), z ∈ U .
Definition 1.4([3]) Let P denote the class of analytic functions p : U → C, p(0) = 1,
and ℜ{p(z)} > 0, then p(z) ≺ 1+z1−z
.
The class P can be completely characterized in terms of subordination. We need the
following lemmas to derive our results.
Lemma 1.5([3]) If the function p ∈ P is given by the series
p(z) = 1 + c1z + c2z2 + c3z
3 + · · · , (1.3)
then the following sharp estimate holds:
|cn| ≤ 2 (n = 1, 2, · · · ).
Lemma 1.6([6]) If the function p ∈ P is given by the series (1.3), then
2c2 = c21 + x(4 − c21), (1.4)
4c3 = c31 + 2(4 − c21)c1x− c1(4 − c21)x2 + 2(4 − c21)(1 − |x|2)z, (1.5)
for some x, z with |x| ≤ 1 and |z| ≤ 1.
§2. Main Results
Definition 2.1 Let ϕ : U → C be analytic, and let the Maclaurin series of ϕ is given by
ϕ(z) = 1 +B1z +B2z2 +B3z
3 + · · · (B1, B2 ∈ R, B1 > 0). (2.1)
Let 0 ≤ γ ≤ 1 and τ ∈ C\{0}. A function f ∈ A is in the class Rτq,γ(ϕ) if it satisfies the
following subordination:
1 +1
τ(∂qf(z) + γz∂2
qf(z) − 1) ≺ ϕ(z).
Some Classes of Analytic Functions with q-Calculus 73
Theorem 2.2 Let 0 ≤ γ ≤ 1, τ ∈ C\{0} and let the function f as in (1.1) be in the class
Rτq,γ(ϕ). Also let p =
[2]q[4]q(1+γ)(1+[3]qγ)([3]q)2(1+[2]qγ)2 .
(1) If B1, B2 and B3 satisfy the conditions 2|B2|(1− p) +B1(1− 2p) ≤ 0, |B1B3 − pB22 | −
pB21 ≤ 0, then the second Hankel determinant satisfies
|a2a4 − a23| ≤
|τ |2B21
([3]q)2(1 + [2]qγ)2.
(2) If B1, B2 and B3 satisfy the conditions 2|B2|(1− p)+B1(1− 2p) ≥ 0, 2|B1B3 − pB22 |−
2(1−p)B1|B2|−B1 ≥ 0, or the conditions 2|B2|(1−p)+B1(1−2p) ≤ 0, |B1B3−pB22 |−pB2
1 ≥0, then the second Hankel determinant satisfies
|a2a4 − a23| ≤
|τ |2[2]q[4]q(1 + γ)(1 + [3]qγ)
|B1B3 − pB22 |.
(3) If B1, B2 and B3 satisfy the conditions 2|B2|(1− p) +B1(1− 2p) > 0, |B1B3 − pB22 | −
pB21 ≤ 0, then the second Hankel determinant satisfies
|a2a4 − a23| ≤
|τ |2B21
4[4]q[2]q(1 + γ)(1 + [3]qγ)
×[4p|B1B3 − pB2
2 | − 4B1(1 − p)[|B2|(3 − 2p) +B1] − 4B22(1 − p)2 −B2
1(1 − 2p)2
|B1B3 − pB22 | −B1(1 − p)(2|B2| +B1)
].
Proof Since f ∈ Rτq,γ(ϕ), there exists an analytic function w with w(0) = 0 and |w(z)| < 1
in U such that
1 +1
τ(∂qf(z) + γz∂2
qf(z) − 1) = ϕ(w(z)). (2.2)
Define the function p1 by
p1(z) =1 + w(z)
1 − w(z)= 1 + c1z + c2z
2 + · · · ,
or equivalently,
w(z) =p1(z) − 1
p1(z) + 1=
1
2
(c1z +
(c2 −
c212
)z2 +
(c3 − c1c2 +
c314
)z3 + · · ·
)(2.3)
Then p1 is analytic in U with p1(0) = 0 and has a positive real part in U . By using (2.3)
together with (2.1), it is evident that
ϕ
(p1(z) − 1
p1(z) + 1
)= 1 +
1
2B1c1z +
(1
2B1
(c2 −
c212
)+
1
4B2c
21
)z2 + · · · (2.4)
74 Hamid Shamsan, Fuad S. M. Alsarari and S. Latha
since f has the Maclaurin series given by (1.1), a computation shows that
1 +1
τ(∂qf(z) + γz∂2
qf(z) − 1) = 1 +[2]qa2(1 + γ)
τz +
[3]qa3(1 + [2]qγ)
τz2
+[4]qa4(1 + [3]qγ)
τz3 + · · · . (2.5)
It follows from (2.2), (2.4) and (2.5) that
a2 =τB1c1
2[2]q(1 + γ),
a3 =τB1
4[3]q(1 + [2]qγ)
[2c2 + c21(
B2
B1− 1)
],
a4 =τ
8[4]q(1 + [3]qγ)
[B1(4c3 − 4c1c2 + c31) + 2B2c1(2c2 − c21) +B3c
31
].
Therefore,
a2a4 − a23
=τ2B1c1
16([4]q[2]q)(1 + γ)(1 + [3]qγ)
[B1(4c3 − 4c1c2 + c31) + 2B2c1(2c2 − c21) +B3c
31
]
− τ2B21
16([3]q)2(1 + [2]qγ)2
[4c22 + c41(
B2
B1− 1)2 + 4c2c
21(B2
B1− 1)
]
=τ2B1c1
16([4]q[2]q)(1 + γ)(1 + [3]qγ)
{[(4c1c3 − 4c2c
21 + c41
)+
2B2c21
B1
(2c2 − c21
)+B3
B1c41
]
− [4]q[2]q(1 + γ)(1 + [3]qγ)
([3]q)2(1 + [2]qγ)2
[4c22 + c41(
B2
B1− 1)2 + 4c2c
21(B2
B1− 1)
]},
which yields
|a2a4 − a23| = T
∣∣∣∣4c1c3 + c41
[1 − 2
B2
B1− p(
B2
B1− 1)2 +
B3
B1
]
−4pc22 − 4c21c2
[1 − B2
B1+ p(
B2
B1− 1)
]∣∣∣∣ , (2.6)
where,
T =|τ |2B2
1
16([4]q[2]q)(1 + γ)(1 + [3]qγ)and p =
[4]q[2]q(1 + γ)(1 + [3]qγ)
([3]q)2(1 + [2]qγ)2.
It can be easily verified that p ∈ [6481 ,89 ] for 0 ≤ γ ≤ 1 and 0 ≤ q ≤ 1. Let
d1 = 4, d2 = −4[1 − B2
B1+ p(B2
B1− 1)
],
d3 = −4p, d4 =[1 − 2B2
B1− p(B2
B1− 1)2 + B3
B1
].
(2.7)
Then (2.6) becomes
∣∣a2a4 − a23
∣∣ = T∣∣d1c1c3 + d2c
21c2 + d3c
22 + d4c
41
∣∣ . (2.8)
Some Classes of Analytic Functions with q-Calculus 75
Since the function p(eiθz)(θ ∈ R) is in the class P for any p ∈ P , there is no loss of
generality in assuming c1 > 0. Write c1 = c, c ∈ [0, 2]. Substituting the values of c2 and c3
respectively from (1.6) and (1.5) in (2.8), we obtain
|a2a4 − a23| =
T
4
∣∣c4 (d1 + 2d2 + d3 + 4d4) + 2xc2(4 − c2) (d1 + d2 + d3)
+ (4 − c2)x2(−d1c
2 + d3(4 − c2))
+ 2d1c(4 − c2)(1 − |x|2z)∣∣ .
Replacing |x| by µ and substituting the values of d1, d2, d3 and d4 from (2.7) yields
|a2a4 − a23| ≤ T
4
[4c4∣∣∣∣B3
B1− p
B22
B21
∣∣∣∣+ 8
∣∣∣∣B2
B1
∣∣∣∣µc2(4 − c2)(1 − p)
+(4 − c2)µ2(4c2 + 4p(4 − c2)) + 8c(4 − c2)(1 − µ2)]
= T
[c4∣∣∣∣B3
B1− p
B22
B21
∣∣∣∣+ 2c(4 − c2) + 2µ
∣∣∣∣B2
B1
∣∣∣∣ c2(4 − c2)(1 − p)
+µ2(4 − c2)(1 − p)(c− α)(c − β)]≡ F (c, µ), (2.9)
where α = 2, β = 2p/(1 − p) > 2.
Note that for (c, µ) ∈ [0, 2]× [0, 1], differentiating F (c, µ) in (2.9) partially with respect to
µ yields∂F
∂µ= T
[2
∣∣∣∣B2
B1
∣∣∣∣ c2(4 − c2)(1 − p) + 2µ(4 − c2)(1 − p)(c− α)(c− β)
]. (2.10)
Then, for 0 < µ < 1, 0 < q < 1 and any fixed c with 0 < c < 2, it is clear from (2.10) that∂F∂µ
> 0, that is, F (c, µ) is an increasing function of µ. Hence, for fixed c ∈ [0, 2], the maximum
of F (c, µ) occurs at µ = 1, and
maxF (c, µ) = F (c, 1) ≡ G(c),
which is
G(c) = T
{c4[∣∣∣∣B3
B1− p
B22
B21
∣∣∣∣− (1 − p)
(2
∣∣∣∣B2
B1
∣∣∣∣+ 1
)]+ 4c2
[2
∣∣∣∣B2
B1
∣∣∣∣ (1 − p) + 1 − 2p
]+ 16p
}.
Let
X =
∣∣∣∣B3
B1− p
B22
B21
∣∣∣∣− (1 − p)
(2
∣∣∣∣B2
B1
∣∣∣∣+ 1
),
Y = 4
[2
∣∣∣∣B2
B1
∣∣∣∣ (1 − p) + 1 − 2p
], (2.11)
Z = 16p.
Since
max(Xt2 + Y t+ Z) =
Z, Y ≤ 0, X ≤ −Y4 ;
16X + 4Y + Z, Y ≥ 0, X ≥ −Y8 orY ≤ 0, X ≥ −Y
4 ;
4XZ−Y 2
4X, Y > 0, X ≤ −y
8 ,
(2.12)
76 Hamid Shamsan, Fuad S. M. Alsarari and S. Latha
where 0 ≤ t ≤ 4. Then we have
|a2a4 − a23| ≤ B1
16([4]q − 1)([3]q − 1)([2]q − 1)
×
Z, Y ≤ 0, X ≤ −Y4 ;
16X + 4Y + Z, Y ≥ 0, X ≥ −Y8 or Y ≤ 0, X ≥ −Y
4 ;
4XZ−Y 2
4X, Y > 0, X ≤ −y
8 ,
where X,Y and Z are given by (2.11). 2Remark 2.3 notice that
(1) As q → 1− Theorem 2.2 reduces to Theorem 3 in [12].
(2) As q → 1− for the choice ϕ(z) := (1 + Az)/(1 + Bz) with −1 ≤ B < A ≤ 1 Theorem
2.2 reduces to Theorem 2.1 in [12].
Definition 2.4 An analytic function f is close-to-q-convex in U , if and only if, there exists
g ∈ S∗q such that
ℜ{z∂qf(z)
g(z)
}> 0.
We denote the class of close-to-q-convex functions by Kq.
For f ∈ S∗, we can write z∂qf(z) = f(z)h(z), where h ∈ P , the class of function satisfying
ℜh(z) > 0 for z ∈ U and
h(z) = 1 +
∞∑
n=2
cnzn.
For f ∈ Kq, we can write z∂qf(z) = g(z)p(z), where p ∈ P and
p(z) = 1 +
∞∑
n=2
pnzn.
Theorem 2.5 Let f ∈ Kq and given by (1.1) with associated starlike function g define by
g(z) = z +
∞∑
n=2
bnzn.
Then
T2(2) = |a23 − a2
2| ≤ [5]q, (b2 ∈ R)
and the inequality is sharp.
Proof Write z∂qf(z) = g(z)h(z) and zg′(z) = g(z)p(z), with
h(z) = 1 +∞∑
n=1
cnzn
Some Classes of Analytic Functions with q-Calculus 77
and
p(z) = 1 +
∞∑
n=1
pnzn.
Then equating the coefficients in z∂qf(z) = g(z)h(z) where coefficients’ relations from zg′(z) =
g(z)p(z) is also used, we obtain
a2 =c1 + p1
[2]q,
a3 =p21 + p2 + 2p1c1 + 2c2
2[3]q
so that
|a23 − a2
2| =
∣∣∣∣−1
[2]2qc21 +
1
[3]2qc22 −
2
[2]2qc1p1 +
2
[3]2qc1c2p1 −
1
[2]2qp21 +
1
[3]2qc21p
21 +
1
[3]2qc2p
21
+1
[3]2qc1p
31 +
1
4[3]2qp41 +
1
[3]2qc2p2 +
1
[3]2qc1p1p2 +
1
2[3]2qp21p2 +
1
4[3]2qp22
∣∣∣∣ .
We now use Lemma 1.6 to express c2 and p2 in terms of c1 and p1 and writingX = 4 − c21and Y = 4 − p2
1 for simplicity to get
|a23 − a2
2|
=
∣∣∣∣−1
[2]2qc21 +
1
4[3]2qc41 −
2
[2]2qc1p1 +
1
[3]2qc31p1 −
1
[2]2qp21 +
7
4[3]2qc21p
21
+3
2[3]2qc1p
31 +
9
16[3]2qp41 +
1
2[3]2qc21xX +
1
[3]2qc1p1xX +
3
4[3]2qp21xX
+1
4[3]2qx2X2 +
1
4[3]2qc21yY +
1
2[3]2qc1p1yY
+3
8[3]2qp21yY +
1
4[3]2qxXyY +
1
16[3]2qy2Y 2
∣∣∣∣ .
Without loss in generality we can assume that c1 = c where 0 ≤ c ≤ 2. Also since we are
assuming b2 = p1 to be real, we can write p1 = r, with 0 ≤ |r| ≤ 2, and write |r| = p. We
note at this point a further normalisation of p1 to be real would remove the requirement that
p1 = b2 is real, but such normalisation does not appear to be justified. It follows from Lemma
1.6 that with now X = 4 − c2 and Y = 4 − p2. So,
|a23 − a2
2|
≤∣∣∣∣−1
[2]2qc2 +
1
4[3]2qc4 − 2
[2]2qcp+
1
[3]2qc3p− 1
[2]2qp2 +
7
4[3]2qc2p2 +
3
2[3]2qcp3 +
9
16[3]2qp4
∣∣∣∣
+1
2[3]2qc2|x|X +
1
[3]2qcp|x|X +
3
4[3]2qp2|x|X +
1
4[3]2q|x|2X2 +
1
4[3]2qc2|y|Y
+1
2[3]2qcp|y|Y +
3
8[3]2qp2|y|Y +
1
4[3]2q|x|X |y|Y +
1
16[3]2q|y|2Y 2.
78 Hamid Shamsan, Fuad S. M. Alsarari and S. Latha
Now we assume |x| ≤ 1 and |y| ≤ 1 and simplify to obtain
|a23 − a2
2| ≤∣∣∣∣−1
[2]2qc2 +
1
4[3]2qc4 − 2
[2]2qcp+
1
[3]2qc3p− 1
[2]2qp2 +
7
4[3]2qc2p2 +
3
2[3]2qcp3 +
9
16[3]2qp4
∣∣∣∣
+9
[3]2q− 1
4[3]2qc4 +
6
[3]2qcp− 1
[3]2qc3p+
3
[3]2qp2 − 3
4[3]2qc2p2 − 1
2[3]2qcp3 − 5
16[3]2qp4.
Suppose that the expression between the modulus signs is positive, then
|a23 − a2
2| ≤ ψ1(c, p) =9
[3]2q− 1
[2]2qc2 +
2(3[2]2q − [3]2q)
[2]2q[3]2qcp
+2(3[2]2q − [3]2q)
[2]2q[3]2qp2 +
1
[3]2qc2P 2 +
1
[3]2qcp3 +
1
4[3]2qp4.
Then for 0 ≤ c ≤ 2 and 0 ≤ p ≤ 2 and fixed q with 0 < q < 1 and calculus we get that ψ1(c, p)
has a maximum value of [5]q at [0,2].
If the expression between the modulus signs is negative, then
|a23 − a2
2| ≤ ψ2(c, p) =9
[3]2q+
1
[2]2qc2 − 1
2[3]2qc4 +
2(3[2]2q + [3]2q)
[2]2q[3]2qcp
− 2
[3]2qc3p
3[2]2q + [3]2q[2]2q[3]2q
p2 − 5
2[3]2qc2P 2 − 2
[3]2qcp3 − 7
8[3]2qp4.
Then for 0 ≤ c ≤ 2 and 0 ≤ p ≤ 2 and fixed q with 0 < q < 1 and calculus we get that ψ2(c, p)
has a maximum value less than [3]q . Thus the proof is complete. 2As q → 1−, we have following result due to D. K. Thomas and S. Abdul Halim [18].
Corollary 2.6 Let f ∈ K and be given by (??) with the associated starlike function g be defined
by
g(z) = z +
∞∑
n=2
bnzn.
Then
T2(2) = |a23 − a2
2| ≤ 5,
provided b2 is real. The inequality is sharp.
References
[1] H. S. Al-Amiri and M. O. Reade, On a linear combination of some expressions in the theory
of the univalent functions, Monatshefte Math., 80(4)(1975), 257-264.
[2] D. Bansal, Upper bound of second Hankel determinant for a new class of analytic functions,
Appl. Math. Lett., 26(1)(2013), 103-107.
[3] P. L. Duren, Univalent Functions. Grundlehren der Mathematischen Wissenschaften,
Springer, New York, 259(1983).
Some Classes of Analytic Functions with q-Calculus 79
[4] M. Fekete and G. Szego , Eine Bermerkung Uber ungerade schlichte Funktionen, J. London
Math. Soc., 1(2)(1933), 85-89.
[5] S. Gagandeep and S. Gurcharanjit, Second Hankel determinant for a Subclass of alpha
Convex functions, J. Appl.Computat Math., 3(2014), 1-3.
[6] U. Grenander, G. Szego, Toeplitz Forms and Their Applications, California Monographs
in Mathematical Sciences. University of California Press, Berkeley, (1958).
[7] A. W. Goodman, Univalent Functions, Vol I, II. Washington, New Jersey: Polygonal
Publishing House, (1983).
[8] W. K. Hayman, On the second Hankel determinant of mean univalent functions, Proc.
London Math. Soc., 3(18)(1968), 77-94.
[9] F. H. Jackson,On q-functions and a certain difference operator, Trans. Royal Soc. Edin-
burgh, 46(1909), 253-281.
[10] A. Janteng, S. Halim and M. Darus, Hankel determinant forstarlike and convex functions,
Int. J. Math. Anal. (Ruse), 1 (2007), 619-625.
[11] Y. Ke and Lek-Heng Lim, Every matrix is a product of Toeplitz matrices, Foundations of
Computational Mathematics, 16(30) (2016), 577-598.
[12] L. S. Keong, V. Ravichandran, and S. Supramaniam. Bounds for the second Hankel de-
terminant of certain univalent functions, arXiv :1303.0134 (2013).
[13] M. S. Liu, J. F. Xu and M. Yang, Upper bound of second Hankel determinant for certain
subclasses of analytic functions, Abstr. Appl. Anal., 2014.
[14] K. I. Noor and S. A. Al-Bany, On Bazilevic functions, Internat. J. Math. Math. Sci.,
1(10)(1987), 79-88.
[15] C. Pommerenke, On the coefficients and Hankel determinants of univalent functions, J.
London Math. Soc., 41(1966), 111-122.
[16] C. Pommerenke, On the Hankel determinants of univalent functions, Math., 14(1967),
108-112.
[17] C. Pommerenke, G. Jensen, Univalent functions, Vandenhoeck und Ruprecht, 25, 1975.
[18] D. K. Thomas, S. A. Halim, Toeplitz matrices whose elements are the coefficients of starlike
and close-to-convex functions, Bulletin of the Malaysian Mathematical Sciences Society,
(2016), 1-10.
International J.Math. Combin. Vol.4(2019), 80-88
On Obic Algebras
Ilojide Emmanuel
(Department of Mathematics, Federal University of Agriculture, Abeokuta 110101, Nigeria)
E-mail: [email protected], [email protected]
Abstract: In this paper, obic algebras are introduced and their properties are investigated.
Homomorphisms and krib maps as well as monics of obic algebras are studied. Properties
of implicative obic algeras are also investigated.
Key Words: Obic algebras, krib maps, monics.
AMS(2010): 20N02, 20N05, 06F35.
§1. Introduction
Algebras of type (2,0) are well known types of algebraic structures. They comprise non-empty
sets, some constant element together with a binary operation. In [1], Kim and Kim introduced
the notion of BE-algebras. Ahn and so, in [2] and [3] introduced the notions of ideals and upper
sets in BE-algebras and investigated related properties. In this paper, a new class of algebras
called obic algebras are introduced. Their properties are investigated. Homomorphisms and
krib maps as well as monics of obic algebras are studied. Moreover, translations in obic algebras
are investigated as well as properties of implicative obic algebras.
Definition 1.1 A non-empty set X together with a binary operation ∗ defined on X is called
a groupoid.
Definition 1.2 A triple (X ; ∗, 0), where X is a non-empty set, ∗ a binary operation on X and 0
a constant element of X is called an obic algebra if the following axioms hold for all x, y, z ∈ X:
(1) x ∗ 0 = x;
(2) [x ∗ (y ∗ z)] ∗ x = x ∗ [y ∗ (z ∗ x)];(3) x ∗ x = 0.
Example 1.1 Consider the multiplicative group G = {1,−1, i,−i}. Define a binary operation
∗ on G by a ∗ b = ab−1. Then (G; ∗, 1) is an obic algebra.
Example 1.2 Let Z denote the set of integers. Then (Z;−, 0) is an obic algebra.
Example 1.3 Let X = {0, 1}. Define a binary operation ∗ on X in Table 1.
1Received June 11, 2019, Accepted December 2, 2019.
On Obic Algebras 81
∗ 0 1
0 0 1
1 1 0
Table 1
Then, (X, ∗, 0) is an obic algebra.
We shall adopt the notation X for an obic algebra (X ; ∗, 0) unless stated otherwise.
Definition 1.3 An obic algebra is called simple if y ∗ (z ∗ x) = x ∗ (y ∗ z) for all x, y, z ∈ X.
Definition 1.4 An obic algebra is called plain if 0 ∗ (y ∗ z) = (0 ∗ y) ∗ z for all y, z ∈ X.
Definition 1.5 An obic algebra X is said to have the weak property (WP) if x ∗ y = 0 and
y ∗ x = 0 imply that x = y.
Definition 1.6 An obic algebra X is called prime if 0 ∗ x = 0 for all x ∈ X.
Lemma 1.1 Let X be an obic algebra. Then for all x, y ∈ X, the following hold:
x ∗ y = [x ∗ (y ∗ x)] ∗ x.
Definition 1.7 A non-empty subset S of an obic algebra X is called a subalgebra if S is an
obic algebra with respect to the binary operation in X.
Example 1.4 Let X be an obic algebra. Then X and {0} are subalgebras of X .
Example 1.5 LetX be the obic algebra in example 1.1. Then the subset {1,−1} is a subalgebra
of X .
The following results are immediately obtained by the definition.
Proposition 1.1 A non-empty subset S of an obic algebra is a subalgebra if and only if the
following hold:
(1) 0 ∈ S;
(2) x ∗ y ∈ S for all x, y ∈ S.
Proposition 1.2 Let X be a plain obic algebra. Then, the subset S = {x ∈ X : 0 ∗ x = 0} is a
subalgebra of X.
§2. Obic Homomorphisms
Definition 2.1 Let (X ; ∗, 0) and (Y ; ◦, 0′) be obic algebras. A function f : X → Y is called
an obic homomorphism if f(a ∗ b) = f(a) ◦ f(b) for all a, b ∈ X.
Definition 2.2 Let f : X → Y be an obic homomorphism. The set {x ∈ X : f(x) = 0′} is
called the kernel of f .
82 Ilojide Emmanuel
Proposition 2.1 Let f : X → Y be an obic homomorphism. Then the kernel of f is a
subalgebra of X.
Then, we get conclusions following by definition.
Proposition 2.2 Let f : X → Y be an obic homomorphism. Then,
(1) f(0) = 0′;
(2) x ∗ y = 0 ⇒ f(x) ◦ f(y) = 0′ for all x, y ∈ X.
Let f : X → Y be an obic homomorphism. Define a relation ∼ by (x ∼ y) ⇔ f(x) = f(y).
Then, we know
Lemma 2.1 Let f : X → Y be an obic homomorphism. The relation ∼ defined by (x ∼ y) ⇔f(x) = f(y) is an equivalence relation.
Definition 2.3 An equivalence relation ∼ on an obic algebra X is called a congruence if (x ∼ y)
and (u ∼ v) ⇒ (x ∗ u) ∼ (y ∗ v).
We have the following result by definition.
Lemma 2.2 Let f : X → Y be an obic homomorphism. The equivalence relation ∼ defined by
(x ∼ y) ⇒ f(x) = f(y) is a congruence.
Let [x] be the equivalence class of x ∈ X and let X denote the collection of equivalence
classes in the equivalence relation ∼. Define a binary operation ⋄ on X by [x] ⋄ [y] = [x ∗ y].
Theorem 2.1 Let f : X → Y be an obic homomorphism. Then (X ; ⋄, [0]) is an obic algebra.
Proof By Lemma 2.2, the binary operation ⋄ is well-defined. Now, let [x], [y], [z] ∈ X.
Consider [x] ⋄ [0] = [x ∗ 0] = [x]. Also,
([x] ⋄ ([y] ⋄ [z])) ⋄ [x] = ([x] ⋄ [y ∗ z]) ⋄ [x] = ([x ∗ (y ∗ z)]) ⋄ [x]
= [(x ∗ (y ∗ z)) ∗ x] = [x ∗ (y ∗ (z ∗ x))]= [x] ⋄ ([y] ⋄ ([z] ⋄ [x])).
Also, [x] ⋄ [x] = [x ∗ x] = [0]. 2Theorem 2.2 Let f : X → X be an endomorphism. Then f(X) is isomorphic to X.
Proof Consider the map φ : f(X) → X such that φ(y) = [y]. Let y1, y2 ∈ f(X). Then
φ(y1 ∗ y2) = [y1 ∗ y2] = [y1] ⋄ [y2] = φ(y1) ⋄ φ(y2). Also, φ is one to one and onto. 2Theorem 2.3 Let φ : X → X be an obic homomorphism; where X has the weak property.
Then φ is one to one if and only if ker(φ) = {0}.
Proof Suppose φ is one to one. Let x ∈ ker(φ). Then φ(x) = 0 = φ(0). So, ker(φ) = {0}.
On Obic Algebras 83
Conversely, suppose ker(φ) = {0}. Let x, y ∈ X such that φ(x) = φ(y). Then φ(x ∗ y) =
φ(x) ∗ φ(y) = 0. Also, φ(y ∗ x) = 0. So, (x ∗ y), (y ∗ x) ∈ ker(φ). Hence φ is one to one. 2Definition 2.4 An obic homomorphism f : X → X is called idempotent if f(f(x)) = f(x) for
all x ∈ X.
Theorem 2.4 Let X be an obic algebra with weak property. Let φ be an idempotent endomor-
phism on X. Then φ is one to one if and only if φ is the identity map.
Proof Suppose φ is one to one. Let x ∈ X . Then φ((x ∗ φ(x))) = φ(x) ∗ φ(φ(x)) =
φ(x) ∗ φ(x) = 0 = φ(0).So, x ∗ φ(x) = 0. Similarly argument gives φ(x) ∗ x = 0. And so
φ(x) = x. Hence φ is the identity map.
The converse is obvious. 2§3. Implicative Obic Algebras
Definition 3.1 An obic algebra X is called implicative if x ∗ (y ∗ x) = x for all x, y ∈ X.
The following conclusion can be obtained by the definition.
Lemma 3.1 Let X be an implicative obic algebra. Then the following hold:
(1) 0 ∗ 0 = 0;
(2) x ∗ y = (x ∗ y) ∗ (0 ∗ y);(3) x ∗ y = (x ∗ (y ∗ x)) ∗ y;(4) y ∗ x = y ∗ (x ∗ (y ∗ x)).
Definition 3.2 Let X be an obic algebra. Let x be a fixed element of X. The map Lx : X → X
such that Lx(a) = x ∗ a for all a ∈ X is called a left translation on X. Similarly, the map
Rx : X → X such that Rx(a) = a ∗ x for all a ∈ X is called a right translation on X.
Theorem 3.1 Let Lx : X → X be an endomorphism. Then x = 0. Moreover, if X is
implicative, then x = x ∗ (x ∗ y).
Proof Consider x = x ∗ 0 = Lx(0) = Lx(0 ∗ 0) = Lx(0) ∗Lx(0) = (x ∗ 0) ∗ (x ∗ 0) = 0. Now
suppose X is implicative. Let y ∈ X . Then x = x ∗ 0 = Lx(0) = Lx(0 ∗ y) = Lx(0) ∗ Lx(y) =
x ∗ (x ∗ y). 2Denote the collection of left translations on an obic algebra X by L(X) and define a binary
operation ⊙ on L(X) by (La ⊙ Lb)(x) = La(x) ∗ Lb(x) for all x ∈ X .
Theorem 3.2 Let X be an implicative obic algebra. Then (L(X);⊙, L0) is an obic algebra.
Proof Let La, Lb, Lc ∈ L(X). For every x ∈ X , consider (La ⊙ L0)(x) = La(x) ∗ L0(x) =
(a ∗ x) ∗ (0 ∗ x) = (a ∗ x) = La(x). So, La ⊙ L0 = La.
84 Ilojide Emmanuel
Also consider
(La ⊙ (Lb ⊙ Lc) ⊙ La)(x) = (a ∗ x) ∗ ((b ∗ x) ∗ ((c ∗ x) ∗ (a ∗ x)))= (La ⊙ (Lb ⊙ (Lc ⊙ La)))(x).
So,
(La ⊙ (Lb ⊙ Lc) ⊙ La) = (La ⊙ (Lb ⊙ (Lc ⊙ La))).
And clearly,
(La ⊙ La)(x) = La(x) ∗ La(x) = (a ∗ x) ∗ (a ∗ x) = 0 = L0(x). 2Corollary 3.1 Let X be a prime obic algebra. Then (L(X);⊙, L0) is an obic algebra.
Corollary 3.2 (L(X);⊙, L0) is prime if and only if X is prime.
We therefore know that
Proposition 3.1 Let X be an obic algebra. Then the translation L0 : X → X commutes with
any endomorphism on X.
Definition 3.3 An obic algebra X is said to have the distributive property if 0 ∗ (x ∗ y) =
(0 ∗ x) ∗ (0 ∗ y) for all x, y ∈ X.
Proposition 3.2 Let X be an obic algebra with distributive property. Then the translation
L0 : X → X is the only homomorphism in the collection L(X).
Proof Clearly, L0 is a homomorphism. Let x ∈ X such that x 6= 0. Let Suppose Lx is a
homomorphism on X . Consider x = (x ∗ 0) = Lx(0) = Lx(0 ∗ 0) = Lx(0) ∗ Lx(0) = 0; which is
a contradiction. 2§4. Krib Maps in Obic Algebras
Definition 4.1 Let X be an obic algebra. A self map α : X → X is called a right krib map if
α(x ∗ y) = x ∗ α(y) for all x, y ∈ X.
If α(x ∗ y) = α(x) ∗ y for all x, y ∈ X, then α is called a left krib map. α is called a krib
map if it is both a right and a left krib map.
Example 4.1 Consider the obic algebra X in Example 1.3. Define α : X → X by α(1) =
0, α(0) = 1. Then α is a right krib map of X .
Denote by D(X) the collection of right krib maps on an obic algebra X . We know the next
result by definition.
Proposition 4.1 Let X be an obic algebra. Then D(X) is a monoid.
On Obic Algebras 85
Definition 4.2 Let X be an obic algebra. A map α : X → X is called regular if α(0) = 0. If
α(0) 6= 0, then α is called irregular.
The following results can be verified immediately.
Lemma 4.1 Let α be an irregular right krib map of an obic algebra X. Then the following
hold for all x ∈ X:
(1) x ∗ α(x) 6= 0;
(2) α(x) = x ∗ α(0).
Proposition 4.2 Every right krib map of a prime obic algebra is regular.
Corollary 4.1 Let α be a krib map on a prime obic algebra X. Then α is regular.
Proposition 4.3 A right krib map α of an obic algebra X is regular if and only if x ∗α(x) = 0
for all x ∈ X.
Proof Suppose α is regular. Then 0 = α(0) = xα(x). Conversely, suppose x ∗ α(x) = 0.
Then α(0) = x ∗ α(x) = 0. 2Proposition 4.4 A left krib map α of an obic algebra X is regular if and only if α(x) ∗ x = 0
for all x ∈ X.
Theorem 4.1 Let α be a krib map of an obic algebra X. Then the following are equivalent:
(1) x ∗ α(x) = 0;
(2) α is regular;
(3) α(x) ∗ x = 0.
Proof The proof is straightforward by definition. 2§5. Monics of Obic Algebras
Let X be an obic algebra. Define ’∧’ by x ∧ y = y ∗ (y ∗ x) for all x, y ∈ X .
Definition 5.1 Let X be an obic algebra. A function θ : X → X is called a left (resp.
right)monic if θ(x ∗ y) = (θ(x) ∗ y) ∧ (x ∗ θ(y)) (resp. θ(x ∗ y) = (x ∗ θ(y)) ∧ (θ(x) ∗ y)) for all
x, y ∈ X.
If θ : X → X is both a left and a right monic, then θ is called a monic.
Example 5.1 Let X be the obic algebra given by Table 2.
∗ 0 1
0 0 1
1 1 0
Table 2
86 Ilojide Emmanuel
The map θ : X → X such that θ(1) = 0, θ(0) = 1 is a left monic.
Definition 5.2 Let X be an obic algebra. A map α : X → X is called regular if α(0) = 0.
Definition 5.3 Let X be an obic algebra. A self map θ on X is called self preserving if
θ(x) ∗ x = x for all x ∈ X.
Definition 5.4 Let X be an obic algebra. A self map θ on X is called anti-self preserving if
x ∗ θ(x) = x for all x ∈ X.
Definition 5.5 Let X be an obic algebra. A self map θ on X is called preserving if it is both
self-preserving and ant-self-preserving.
Proposition 5.1 Let θ be a regular left monic on an obic algebra X. Then,
(x ∗ θ(x)) ∗ [(x ∗ θ(x)) ∗ (θ(x) ∗ x)] = (y ∗ θ(y)) ∗ [(y ∗ θ(y)) ∗ (θ(y) ∗ y)]
for all x, y ∈ X.
Proof Now, 0 = θ(0) = θ(x ∗ x) = (x ∗ θ(x)) ∗ [(x ∗ θ(x)) ∗ (θ(x) ∗ x)]. Similar argument
gives also (y ∗ θ(y)) ∗ [(y ∗ θ(y)) ∗ (θ(y) ∗ y)] = 0. Hence, the conclusion follows. 2Proposition 5.2 Let X be a regular left monic on an associative obic algebra X. Then
0 ∗ [θ(x) ∗ x] = 0 ∗ [θ(y) ∗ y] for all x, y ∈ X.
Proof By proposition 5.1,
0 = [x ∗ θ(x)] ∗ [(x ∗ θ(x)) ∗ (θ(x) ∗ x)]= [(x ∗ θ(x)) ∗ (x ∗ θ(x))] ∗ [θ(x) ∗ x]= 0 ∗ [θ(x) ∗ x].
Similarly, we have 0 ∗ [θ(y) ∗ y] = 0. The conclusion follows. 2Proposition 5.3 Let θ be a self preserving left monic on an obic algebra X. Then
[x ∗ θ(x)] ∗ [(x ∗ θ(x)) ∗ x)] = θ(0) for all x ∈ X.
Proof Now,
θ(0) = θ(x ∗ x)= [θ(x) ∗ x] ∧ [x ∗ θ(x)]= [x ∗ θ(x))] ∗ [(x ∗ θ(x)) ∗ x]. 2
Corollary 5.1 Let θ be a regular self preserving left monic on an obic algebra X.
Then [x ∗ θ(x))] ∗ [(x ∗ θ(x)) ∗ x] = 0 for all x ∈ X.
The following propositions can be immediately verified.
On Obic Algebras 87
Proposition 5.4 Let θ be a regular self preserving left monic on an obic algebra X. Then
[x ∗ θ(x))] ∗ [(x ∗ θ(x)) ∗ x] = [y ∗ θ(y))] ∗ [(y ∗ θ(y)) ∗ y] for all x, y ∈ X.
Proposition 5.5 Let θ be a self preserving left monic on an associative obic algebra X. Then
0 ∗ x = θ(0) for all x ∈ X.
Proposition 5.6 Let θ be a regular self preserving left monic on an associative obic algebra
X. Then 0 ∗ [θ(x) ∗ x] = 0 ∗ [θ(y) ∗ y] for all x, y ∈ X.
Proposition 5.7 Let θ be a regular self preserving left monic on an associative obic algebra
X. Then 0 ∗ x = 0 for all x ∈ X.
Theorem 5.1 Let X be an associative obic algebra with a self preserving left monic θ. Then
X is prime if and only if θ is regular.
Proof Suppose X is prime. Then,
0 = 0 ∗ x = [(x ∗ θ(x)) ∗ (x ∗ θ(x))] ∗ x= (x ∗ θ(x)) ∗ [(x ∗ θ(x)) ∗ x]= x ∧ [x ∗ θ(x)]= θ(x ∗ x) = θ(0).
Conversely, suppose θ is regular. Then,
0 = θ(0) = θ(x ∗ x) = [(θ(x) ∗ x)] ∧ [x ∗ θ(x)]= x ∧ [x ∗ θ(x)] = 0 ∗ x. 2
The two conclusions following can be easily verified by definition.
Proposition 5.8 Let θ be an anti-self preserving left monic on an obic algebra X. Then
x ∗ [x ∗ θ(x)] = θ(0) for all x ∈ X.
Proposition 5.9 Let θ be a regular anti-self preserving left monic on an obic algebra X. Then
y ∗ [y ∗ θ(y)] = x ∗ [x ∗ θ(x)] for all x, y ∈ X.
Theorem 5.2 Let θ be an anti-self preserving left monic on an associative obic algebra X.
Then 0 ∗ [θ(x) ∗ x] = θ(0). Moreover, if θ is regular, then 0 ∗ [θ(x) ∗ x] = 0 for all x ∈ X.
Proof Notice that
θ(0) = θ(x ∗ x) = [θ(x) ∗ x] ∧ [x ∗ θ(x)]= [θ(x) ∗ x] ∧ x= 0 ∗ [θ(x) ∗ x].
The second part of the theorem is obvious. 2
88 Ilojide Emmanuel
Proposition 5.10 Let X be an obic algebra with a left monic θ. Then θ(x) = x ∗ [x ∗ θ(x)] for
all x ∈ X.
Proof Notice that
θ(x) = θ(x ∗ 0) = θ(x)
= [θ(x) ∗ 0] ∧ [x ∗ θ(0)]
= x ∗ [x ∗ θ(x)].
This completes the proof. 2Corollary 5.2 Let θ be an anti-self preserving regular left monic on an obic algebra X. Then
θ(x) = 0 for all x ∈ X.
Corollary 5.3 Let θ be a regular left monic on an associative obic algebra X. Then θ(x) =
0 ∗ θ(x) for all x ∈ X.
References
[1] H. S. Kim and Y. H. Kim, On BE-algebras, Sci. Math. Jpn., 66(2007),113–116.
[2] S.S. Ahn and K. S. So, On ideals and upper sets in BE-algebras, Sci. Math. Jpn., 68(2008),
351–357.
[3] S.S. Ahn and K. S. So, On generalized upper sets in BE-algebras, Bull. Korean Math.
Soc., 46(2009), 281–287.
[4] R. H. Bruck, A Survey of Binary Systems, Springer-Verlag, Berlin-Gottingen-Heidelberg,
1966, 185pp.
[5] J. Dene and A. D. Keedwell, Latin Squares and Their Applications, the English University
press Ltd, 1974, 549pp.
International J.Math. Combin. Vol.4(2019), 89-92
Prime BI-Ideals of Po-Γ-Groupoids
J.Catherine Grace John and B. Elavarasan
(Department of Mathematics, Karunya University, Coimbatore-641114, India)
E-mail: [email protected], [email protected]
Abstract: In this paper, we have studied the notion of prime bi-ideals and semi prime
bi-ideals of Po-Γ-groupoids and explored the various properties of prime bi-ideals in Po-Γ-
groupoids. Also we obtained the condition for Po-Γ-groupoids to be regular.
Key Words: Partially ordered Γ-groupoid (po-Γ-groupoid), left (right, two-sided) ideal,
quasi-ideal, bi-ideal, prime (semi prime) bi-ideal, regular partially ordered -Γ-groupoid.
AMS(2010): 06F05, 20N02.
§1. Introduction
In 1983, A.P.J. van der Walt [5] introduced the interesting concepts of prime and semi prime
bi-ideals for an associative ring with unity. In 1995, using the concepts defined by A. P. J.
van der Walt, the structure of a ring containing prime and semi prime bi-ideals were studied
by H. J. le Roux [2]. In 2001, Kehayopulu and Tsingelis [1] studied prime ideals of groupoids.
Following [1], in 2005, S.K. Lee developed prime left (right) ideals of groupoids [3] and obtained
some results on prime bi-ideals of groupoids [4]. In this paper we have studied the notion of
prime bi-ideals and semi prime bi-ideals of Po-Γ-groupoids. Let M be a non empty set. M is
called Γ-groupoid if for all a, b ∈M and γ ∈ Γ, aγb ∈M .
A set (G,Γ,≤) is called a partial order-Γ-groupoid(or simply Po-Γ-groupoid) if
(i) (G,≤) is a partial ordered set;
(ii) (G,Γ) is a Γ-groupoid such that a ≤ b ⇒ aγx ≤ bγx and xγ1a ≤ xγ1b for all a, b, x ∈G; γ, γ1 ∈ Γ.
Throughout this paper G denotes a Po-Γ-groupiod.
A non empty subset A of G is called right(resp.left) ideal of G if
(i) AΓG ⊆ A ( resp.GΓA ⊆ A);
(ii) a ∈ A , b ≤ a for b ∈ G implies b ∈ A.
A non empty subset A is called an ideal of G if it is a right and left ideal of G.
For non-empty subsets A and B of a po-Γ-groupoid G, the product AΓB of A and B
and the subset (A] of G are defined by AΓB = {aγb ∈ S : a ∈ A, b ∈ B, γ ∈ Γ}; (A]=
{x ∈ G : ∃a ∈ A(x ≤ a)}.1Received July 31, 2019, Accepted December 3, 2019.
90 J.Catherine Grace John and B. Elavarasan
A non empty subset Q of G is called a quasi ideal if
(i) (QΓG] ∩ (GΓQ] ⊆ Q;
(ii) a ≤ q; q ∈ Q implies a ∈ Q.
A non empty subset B of G is called a bi-ideal if
(i) (BΓGΓB] ⊆ B;
(ii) a ≤ b; b ∈ B implies a ∈ B.
Every quasi ideal is a bi-ideal. But the converse need not be true. A bi-ideal B of G is prime,
for x, y ∈ G, (xΓGΓy] ⊆ B implies x ∈ B or y ∈ B. A bi-ideal B of G is semi-prime, for
x ∈ G, (xΓGΓx] ⊆ B implies x ∈ B. A non-empty subset I of G is prime if I is an ideal
of G such that for any ideals A,B of G , AB ⊆ I implies A ⊆ I or B ⊆ I. It is clear that
(x)l = (x ∪GΓx](resp.(x)r = (x ∪ xΓG]) is the principle left(resp.right) ideal generated by x.
§2. Main Results
Theorem 2.1 A bi-ideal B of G is prime if and only if for a right ideal R and a left ideal L
of G (RΓL] ⊆ B implies R ⊆ B or L ⊆ B.
Proof Suppose that (RΓL] ⊆ B for a right ideal R and a left ideal L of G and R * B.
Then there exists x ∈ R\B such that (xΓGΓy] ⊆ (RΓGΓL] ⊆ (RΓL] ⊆ B for any y ∈ L which
implies y ∈ B. So L ⊆ B.
Conversely, let (xΓGΓy] ⊆ B for x, y ∈ G. Then (xΓG]Γ(GΓy] ⊆ (xΓGΓGΓy] ⊆ (xΓGΓy] ⊆B. By hypothesis, we have (xΓG] ⊆ B or GΓy] ⊆ B. If (xΓG] ⊆ B, then xΓx ∈ (xΓG]ΓG ⊆ B.
Now, (x)r(x)l = (x∪xΓG]Γ(x∪GΓx] = (xΓx∪xΓGΓx∪xΓGΓx∪xΓGΓGΓx] ⊆ (xΓx∪xΓG] ⊆ B
which implies (x)r ⊆ B or (x)l ⊆ B. Therefore x ∈ B. If (GΓy] ⊆ B, then by the similar
method y ∈ B. 2Theorem 2.2 If a bi-ideal B of G is prime, then B is a left or right ideal of G.
Proof Let B be a prime bi-ideal of G.Then (BΓG] ⊆ B or (GΓB] ⊆ B as (BΓG]Γ(GΓB] ⊆(BΓGΓB] ⊆ B and by Theorem 2.1. So B is a a left ideal or right ideal of G. 2Theorem 2.3 Let G be a po-Γ-groupoid. Then the following statements are hold:
(i) Any left/right/both sided ideal of G is a bi-ideal of G;
(ii) Intersection of right and left ideals of G is a bi-ideal of G;
(iii) Arbitrary intersection of bi-ideals of G is also a bi-ideal of G;
(iv) If B is a bi-ideal of G, then BΓr and rΓB are bi-ideals of G, for any r ∈ G.
Proof This result can be immediately verified by definition. 2Notation 1 For a bi-ideal of B of G, we define LB = {x ∈ B : GΓx ⊆ B}, RB = {x ∈B : xΓG ⊆ B}, IL = {y ∈ LB : yΓG ⊆ LB} and IR = {y ∈ RB : GΓy ⊆ RB}.
Prime BI-Ideals of Po-Γ-Groupoids 91
Theorem 2.4 Let B be bi-ideal of G . Then LB is a left ideal of G contained in B if LB is
non empty.
Proof Let x ∈ LB. Let g ∈ G and γ ∈ Γ . Then gγx ∈ GΓx ⊆ B . Now GΓgγx ⊆GΓGΓx ⊆ GΓx ⊆ B which implies gγx ∈ LB. GΓLB ⊆ LB. hence LB is a left ideal. 2Theorem 2.5 Let B be bi-ideal of G. Then IL is the largest ideal of G contained in B if IL is
non empty. Furthermore, IL coincides with IR.
Proof Let x ∈ IL.Then xΓG ⊆ LB. For any g ∈ G and γ ∈ Γ, we have xγg ∈ xΓG ⊆ LB
and xγgΓG ⊆ xΓGΓG ⊆ xΓG ⊆ LB, So IL is a right ideal of G.
Since IL ⊆ LB ⊆ B, we have x ∈ LB which implies xγg ∈ IL and GΓx ⊆ B.
Now, GΓgγx ⊆ GΓGΓx ⊆ GΓx ⊆ B. So gγx ∈ LB. By Theorem 2.4 and x ∈ IL, we have
xΓG ⊆ LB. Then gγxΓG ⊆ GΓLB ⊆ LB, and we have gγx ∈ IL. Therefore IL is a left ideal.
Let A be an ideal of G such that A ⊆ B. If x ∈ A, then x ∈ B and GΓ ⊆ x ⊆ A ⊆ B
which implies x ∈ LB and A ⊆ LB.
Let x ∈ A. Then xΓG ⊆ A ⊆ LB. Hence x ∈ IL and A ⊆ IL which implies IL is the
largest ideal of G contained in B. Similarly IR is the largest ideal of G contained in B. 2Notation 2 We denote IB as IB :≡ IR = IL by Theorem 2.5.
Theorem 2.6 If B is a prime bi-ideal of G, then IB is a prime ideal of G contained in B.
Proof Let B be a prime bi-ideal of G. Then by Theorem 2.5, IB is an ideal of G.
Suppose XΓY ⊆ IB for any ideals X,Y of G. Since IB ⊆ LB ⊆ B, we have XΓY ⊆ B.
By Theorem 2.1, X ⊆ B or Y ⊆ B. But IB is the largest ideal contained in B, so X ⊆ IB or
Y ⊆ IB which implies IB is a prime ideal of G. 2Corollary 2.7 If B be a semi-prime bi-ideal of G, then IB is a semi-prime ideal of G if IB is
non empty.
Theorem 2.8 If a bi-ideals B of G is semi-prime, then
(i) for any left ideal L of G, LΓL ⊆ B implies L ⊆ B;
(ii) for any right ideal R of G, RΓR ⊆ B implies R ⊆ B.
Proof Suppose LΓL ⊆ B for a left ideal L of G and L * B. Then there exists x ∈ L\B,xΓGΓx ⊆ LΓGΓL ⊆ LΓL ⊆ B. Since B is a semi-prime we have x ∈ B, a contradiction.
The second assertion can be proved similarly. 2Theorem 2.9 If a bi-ideal B of G is semi-prime, then B is a quasi-ideal of G.
Proof Let y ∈ (BΓG] ∩ (GΓB]. Then (yΓGΓy] ⊆ ((BΓG]ΓGΓ(GΓB]] ⊆ (BΓGΓB] ⊆ B.
Since B is a semi prime, we have y ∈ B. Hence B is a quasi-ideal of G. 2Remark 2.10 For a Po-Γ-groupoid G,
(a) The set of all prime ideal of G is denoted by spec(G);
92 J.Catherine Grace John and B. Elavarasan
(b) Bspec(G) denotes the set of prime bi-ideals of G;
(c) Sspec(G) denotes the set of all semi prime bi-ideals of G.
Then we know conclusions following easily by definition.
Theorem 2.11 If G is finite, then⋂spec(G) =
⋂Bspec(G).
Theorem 2.12 A bi-ideal B of G is semi prime if and only if for a right ideal ( left ideal) A
of G (AΓA] ⊆ B implies A ⊆ B.
Theorem 2.13 The intersection of any family of prime bi-ideals of G is a semi prime bi-ideal
of G.
Theorem 2.14 If G is finite, then⋂spec(G) =
⋂Sspec(G).
We note that G is regular if for any x ∈ G, there exist a ∈ G and γ1, γ2 ∈ Γ such that
x ≤ xγ1aγ2x.
The following results shows the necessary and sufficient condition for a Po-Γ-groupiod to
be regular.
Theorem 2.15 Let G be Po-gamma-groupiod. Then G is regular if and only if every bi-ideal
of G is semi-prime.
Proof Let G be regular and B a bi-ideal of G. Suppose that xΓGΓx ⊆ B for x ∈ G. Then
there exist a ∈ G and γ1, γ2 ∈ Γ such that x ≤ xγ1aγ2x ∈ xΓaΓx ∈ xΓGΓx ⊆ B which implies
x ∈ B. Hence B is semi prime.
Conversely , assume that every bi-ideal of G is semi-prime. Let B = (aΓGΓa] for a ∈ G.
Then BΓGΓB = (aΓGΓa]ΓGΓ(aΓGΓa] ⊆ (aΓGΓa] = B, which implies B is a bi-ideal of G and
by assumption (aΓGΓa] is semi-prime. Since aΓGΓa ⊆ (aΓGΓa] = B, we get a ∈ (aΓGΓa] = B.
Then there exist x ∈ G and γ1, γ2 ∈ Γ such that a ≤ aγ1xγ2a. 2References
[1] N. Kehayopulu, On prime ideals of groupoids-ordered groupoids, Scientiae Mathematicae
Japonicae, 5 (2001), 83-86.
[2] H. J. le Roux, A note on prime and semiprime bi-ideals of rings, Kyungpook Math. J., 35
(1995), 243-247.
[3] S.K.Lee, On prime left (right) ideals of grou[iogs- ordered groupiods, Kangweon-Kyungki
Math. J., 13(1) (2005), 13-18.
[4] S.K.Lee, Prime bi-ideals groupiods, Kangweon- Kyungki Math. J., 13(2) (2005), 217-221.
[5] A. P. J. van der Walt, Prime and semiprime bi-ideals,Quaestiones Matematicae, 5 (1983),
341-345.
International J.Math. Combin. Vol.4(2019), 93-101
4-Total Prime Cordiality of Certain Subdivided Graphs
R. Ponraj
(Department of Mathematics, Sri Paramakalyani College, Alwarkurichi-627412, India)
J. Maruthamani and R. Kala
(Department of Mathematics, Manonmaniam Sundarnar University, Tirunelveli-627012, Tamilnadu, India)
E-mail: [email protected], [email protected],[email protected]
Abstract: Let G be a (p, q) graph. Let f : V (G) → {1, 2, · · · , k} be a map where k ∈ N
and k > 1. For each edge uv, assign the label gcd(f(u), f(v)). f is called k-total prime
cordial labeling of G if |tf (i) − tf (j)| ≤ 1, i, j ∈ {1, 2, · · · , k} where tf (x) denotes the total
number of vertices and the edges labelled with x. A graph with a k-total prime cordial
labeling is called k-total prime cordial graph. In this paper we investigate the 4-total prime
cordial labeling for some subdivided graphs.
Key Words: Corona, ladder, triangular snake, k-total prime cordial labeling, Smarandache
k-total prime cordial labeling.
AMS(2010): 05C78.
§1. Introduction
In this paper we consider simple, finite and undirected graphs only. The notion of k- total
prime cordial labelling has been introduced in [4]. In [4–9], they investigate the k-total prime
cordial labeling of some graphs and investigate the 4-total prime cordial labeling behaviour of
path, cycle, star, bistar, ladder, triangular snake, friendship graph, comb, double comb, double
triangular snake, flower graph, gear graph, Jelly fish, book, irregular triangular snake, prism,
helm, dumbbell graph, sunflower graph, dragon, mobius ladder and subdivision of some graphs.
In this paper we examine the 4-total prime cordial labeling of subdivision of some graphs like
star, bistar, comb, double comb, ladder, triangular snake and double triangular snake. Terms
are not defined here follows from [1], [3].
§2. Preliminary Results
Definition 2.1 Let G1, G2 respectively be (p1, q1), (p2, q2) graphs. The corona of G1 with G2,
G1 ⊙G2 is the graph obtained by taking one copy of G1 and p1 copies of G2 and joining the ith
vertex of G1 with an edge to every vertex in the ith copy of G2.
1Received April 14, 2019, Accepted December 5, 2019.
94 R. Ponraj, J. Maruthamani and R. Kala
Definition 2.2 If e = uv is an edge of G then e is said to be subdivided when it is replaced by
the edges uw and wv. The graph obtained by subdividing each edge of a graph G is called the
subdivision graph of G and is denoted by S(G).
§3. k-Total Prime Cordial Labeling
Definition 3.1 Let G be a (p, q) graph. Let f : V (G) → {1, 2, · · · , k} be a function where
k ∈ N and k > 1. For each edge uv, assign the label gcd(f(u), f(v)). f is called k-total
prime cordial labeling of G if |tf (i) − tf (j)| ≤ 1, i, j ∈ {1, 2, · · · , k} where tf (x) denotes the
total number of vertices and the edges labelled with x. A graph with a k-total prime cordial
labeling is called k-total prime cordial graph. Generally, if there are integers i, j ∈ {1, 2, · · · , k}such that |tf (i) − tf (j)| > 1, f is called a Smarandache k-total prime cordial labeling and G a
Smarandache k-total prime cordial labeling graph.
Theorem 3.2 The subdivision of comb S(Pn ⊙K1) is 4-total prime cordial.
Proof Let Pn be the path u1u2 · · ·un. Let xi be the vertex which subdivide the edge
uiui+1. Let vi be the vertex adjacent to ui. Let wi be the pendent vertices vi. Clearly
|V (S(Pn ⊙K1))| + |E(S(Pn ⊙K1))| = 8n− 3.
Case 1. n ≡ 0 (mod 4).
Let n = 4t, t ∈ N. Assign the label 4 to the vertices u1, u2, · · · , ut and assign the label 3 to
the vertices ut+1, ut+2, · · · , u2t. Next we assign the label 2 to the vertices u2t+1, u2t+2, · · · , u3t
and assign 1 to the vertices u3t+1, u3t+2, · · · , un. Now we consider the vertices xi (1 ≤ i ≤n − 1). Assign the label 4 to the vertices x1, x2, · · · , xt and assign the label 3 to the vertices
xt+1, xt+2, . . . , x2t. Next we assign the label 2 to the vertices x2t+1, x2t+2, · · · , x3t. Finally we
assign 1 to the vertices x3t+1, x3t+2, · · · , xn−1. Now we move to the vertices vi, wi (1 ≤ i ≤ n).
Assign the label 4 to the vertices v1, v2, . . . , vt and w1, w2, · · · , wt. Then assign the label 3
to the vertices vt+1, vt+2, · · · , v2t and wt+1, wt+2, · · · , w2t. Now we assign the label 2 to the
vertices v2t+1, v2t+2, · · · , v3t and w2t+1, w2t+1, · · · , w3t. Finally we assign the label 1 to the
vertices v3t+1, v3t+2, · · · , vn and w3t+1, w3t+2, · · · , wn.
Case 2. n ≡ 1 (mod 4).
Let n = 4t+ 1, t ∈ N. As in case 1, assign the label to the vertices ui (1 ≤ i ≤ n− 2), xi
(1 ≤ i ≤ n − 2), vi, wi (1 ≤ i ≤ n − 1). Now we assign the labels 3, 4, 2, 3, 4 respectively to
the vertices un−1, un, xn−1, vn and wn.
Case 3. n ≡ 2 (mod 4).
Let n = 4t + 2, t ∈ N. Assign the label to the vertices ui, vi, wi (1 ≤ i ≤ n − 2) and xi
(1 ≤ i ≤ n− 3) by case 1. Next we assign the labels 4, 3, 4, 3, 2, 2, 4, 3 to the vertices un−1,
un, xn−2, xn−1, vn−1, vn, wn−1 and wn respectively.
Case 4. n ≡ 3 (mod 4).
4-Total Prime Cordiality of Certain Subdivided Graphs 95
Let n = 4t+3, t ∈ N. As in Case 3, assign the label to the vertices ui, vi, wi (1 ≤ i ≤ n−2)
and xi (1 ≤ i ≤ n− 3). Next we assign the labels 2, 3, 4, 3, 3, 3, 2,1 respectively to the vertices
un−1, un, xn−2, xn−1, vn−1, vn, wn−1 and wn. 2Theorem 3.3 The subdivision of double comb S(Pn ⊙ 2K1) is 4-total prime cordial.
Proof Let Pn be the path u1u2 · · ·un. Let zi be the vertex which subdivide the edge
uiui+1. Let vi, wi be the vertices adjacent to ui. Let xi, yi be the pendent vertices adjacent to
ui,vi respectively. Obviously |V (S(Pn ⊙ 2K1))| + |E(S(Pn ⊙ 2K1))| = 12n− 3.
Case 1. n ≡ 0 (mod 4).
Let n = 4t, t ∈ N. Assign the label 4 to the vertices u1, u2, · · · , ut and assign the label 3 to
the vertices ut+1, ut+2, · · · , u2t. Now we assign the label 2 to the vertices u2t+1, u2t+2, · · · , u3t.
Next we assign 1 to the vertices u3t+1, u3t+2, · · · , un−1. Finally we assign the label 3 to the
vertex un. Now we consider the vertices zi (1 ≤ i ≤ n − 1). Assign the label 4 to the vertices
z1, z2, · · · , zt−1 and assign the label 3 to the vertices zt+1, zt+2, · · · , z2t−1. Next we assign
the label 2 to the vertices z2t+1, z2t+2, · · · , z3t−1. Now we assign the label 1 to the vertices
zt, z2t, z3t, z3t+1, · · · , zn−1. Assign the label to the vertices vi, wi, xi, yi (1 ≤ i ≤ n−1). Finally
we assign 2, 4, 4, 3 respectively to the vertices xn, vn, wn and yn.
Case 2. n ≡ 1 (mod 4).
Let n = 4t + 1, t ∈ N. As in case 1, assign the label to the vertices ui, vi, xi, wi, yi
(1 ≤ i ≤ n− 1) and zi (1 ≤ i ≤ n− 2). Now we assign the labels 2, 4, 4, 2, 3, 3 respectively to
the vertices zn−1, xn, vn, un, wn and yn.
Case 3. n ≡ 2 (mod 4).
Let n = 4t+ 2, t ∈ N. Assign the label to the vertices ui, vi, xi, wi, yi (1 ≤ i ≤ n− 1) and
zi (1 ≤ i ≤ n − 2) by case 2. Next we assign the label 1 to zn−1 and assign the labels 4, 4, 2,
3, 3 to the vertices xn, vn, un, wn and yn respectively.
Case 4. n ≡ 3 (mod 4).
Let n = 4t + 3, t ∈ N. As in Case 3, assign the label to the vertices ui, vi, xi, wi, yi
(1 ≤ i ≤ n− 1) and zi (1 ≤ i ≤ n− 2). Next we assign the labels 2, 4, 4, 2, 3, 3 respectively to
the vertices zn−1, xn, vn, un, wn and yn. 2Theorem 3.4 The subdivision of star S(K1,n) is 4-total prime cordial.
Proof Let u be the vertex of degree n and u1, u2, · · · , un be the vertices of degree 2. Let
v1, v2, · · · , vn be the pendent vertices.
Case 1. n ≡ 0 (mod 4).
Let n = 4t, t ∈ N. Assign the label 4 to the vertex u. Next we now move to the vertices
u1, u2, · · · , un. Assign the label 4 to the vertices u1, u2, · · · , ut and assign the label 2 to the
vertices ut+1, ut+2, · · · , u2t. Next we assign the label 3 to the vertices u2t+1, u2t+2, · · · , u3t.
Finally assign the label 1 to the non-labelled vertices of un. Now we consider the pendent
96 R. Ponraj, J. Maruthamani and R. Kala
vertices v1, v2, · · · , vn. Assign the label 4 to the vertices v1, v2, · · · , vt and assign the label 2 to
the vertices vt+1, vt+2, . . . , v2t. Finally assign 3 to the non-labelled vertices of vn.
Case 2. n ≡ 1 (mod 4).
Let n = 4t + 1, t ∈ N. In this case, assign the label to the vertices u, ui (1 ≤ i ≤ n) and
vi (1 ≤ i ≤ n − 2) by in case 1. Next assign the labels 2 and 4 to the vertices vn−1 and vn
respectively.
Case 3. n ≡ 2 (mod 4).
Let n = 4t+ 2, t ∈ N. As in Case 1, assign the label to the vertices u, ui (1 ≤ i ≤ n− 1)
and vi (1 ≤ i ≤ n− 2). Next assign the labels 2, 1 and 4 to the vertices respectively un, vn−1
and vn.
Case 4. n ≡ 3 (mod 4).
Let n = 4t + 3, t ∈ N. Assign the label to the vertices u, ui (1 ≤ i ≤ n − 2) and vi
(1 ≤ i ≤ n − 2) as in case 1. Finally assign the labels 3, 2, 4 and 4 to the vertices un−1, un,
vn−1 and vn respectively. 2Theorem 3.5 The subdivision of bistar S(Bn,n) is 4-total prime cordial.
Proof Let u, v be the vertices of degree n and w be the vertex of degree 2 adjacent to both
u and v. Let ui be the vertex of degree 2 adjacent to u and vi be the vertex of degree 2 adjacent
to v. Let xi and yi (1 ≤ i ≤ n) be the pendent vertex adjacent to ui and vi respectively.
Case 1. n ≡ 0 (mod 4).
Let n = 4t, t ∈ N. Assign the labels 4, 2 and 3 to the vertex u, w and v respectively. Next
we move to the vertices u1, u2, · · · , un. Assign the label 4 to the vertices u1, u2, · · · , u2t and
assign the label 2 to the vertices u2t+1, u2t+2, · · · , u4t. Now we consider the pendant vertices
of un. Assign the label 4 to the vertices x1, x2, · · · , x2t and assign the label 2 to the vertices
x2t+1, x2t+2, · · · , x4t. Now we consider the vertices v1, v2, · · · , vn. Assign the label 3 to the
vertices v1, v2, · · · , v2t and assign the label 1 to the vertices v2t+1, v2t+2, · · · , v4t. Finally we
move to the pendant vertices of vn. Assign the label 3 to the vertices y1, y2, · · · , y2t and assign
the label 1 to the vertices y2t+1, y2t+2, · · · , y4t.
Case 2. n ≡ 1 (mod 4).
Let n = 4t+1, t ∈ N. In this case assign the label to the vertices u, v, w, ui (1 ≤ i ≤ n−1),
vi (1 ≤ i ≤ n− 1), xi (1 ≤ i ≤ n− 1) and yi (1 ≤ i ≤ n− 1) as in case 1. Next assign the labels
4, 2, 3 and 1 respectively to the vertices un, xn, vn and yn.
Case 3. n ≡ 2, 3 (mod 4).
Let n = 4t+ 1 and n = 4t+ 2 t ∈ N. The proof is similar to that of Case 2. 2Theorem 3.6 The subdivision of triangular snake S(Tn) is 4-total prime cordial.
Proof Let Pn be the path u1u2 · · ·un. Let wi be the vertex adjacent to ui and ui+1. Let
4-Total Prime Cordiality of Certain Subdivided Graphs 97
vi be the vertices which subdivide the edge uiui+1 and xi,yi be the vertex which subdivided
uiwi and ui+1wi respectively. It is easy to verify that |V (S(Tn))| + |E(S(Tn))| = 11n− 10.
Case 1. n ≡ 0 (mod 4).
Let n = 4t, t ∈ N. Assign the label 4 to the vertices u1, u2, · · · , ut and assign the label 2 to
the vertices ut+1, ut+2, · · · , u2t. Next we assign the label 3 to the vertices u2t+1, u2t+2, · · · , u3t
then we assign the label 1 to the vertices u3t+1, u3t+2, · · · , un−1. Finally, we assign 3 to the
vertex un. Assign the label 4 to the vertices v1, v2, · · · , vt and assign the label 2 to the vertices
vt+1, vt+2, · · · , v2t−1 and assign the label 3 to the vertices v2t, v2t+1, · · · , v3t−1 then we assign
the label 1 to the vertices v3t, v3t+1, · · · , vn−1. Assign the label to the vertices xi (1 ≤ i ≤ n−1)
as in vi (1 ≤ i ≤ n − 1). Now relabel the vertex x2t by 2. Assign the label 4 to the vertices
y1, y2, · · · , yt−1 and assign the label 2 to the vertices yt, yt+1, · · · , y2t−1 and assign the label 3 to
the vertices y2t, y2t+1, · · · , y3t−1 then we assign the label 1 to the vertices y3t, y3t+1, · · · , yn−2.
Finally we assign the label 2 to the vertices yn−1. Now we consider the vertices wi (1 ≤ i ≤n− 1). Assign the label 4 to the vertices w1, w2, · · · , wt and assign the label 2 to the vertices
wt+1, wt+2, · · · , w2t−1 and assign the label 3 to the vertices w2t, w2t+1, · · · , w3t−1 then we assign
the label 1 to the vertices w3t, w3t+1, · · · , wn−2. Finally we assign 4 to the vertex wn−1.
Case 2. n ≡ 1 (mod 4).
Let n = 4t+ 1, t ∈ N. As in Case 1, assign the label to the vertices ui (1 ≤ i ≤ n− 1), vi,
xi, yi, wi (1 ≤ i ≤ n− 2). Next we assign the labels 3, 3, 2, 4, 4 to the vertices vn−1, un, xn−1,
yn−1 and wn−1 respectively.
Case 3. n ≡ 2 (mod 4).
Let n = 4t + 2, t ∈ N. Assign the label to the vertices ui (1 ≤ i ≤ n − 3), vi, xi, wi
(1 ≤ i ≤ n−3) and yi (1 ≤ i ≤ n−4) as in Case 2. Now we assign the labels 4, 3, 3 respectively
to the vertices un−2, un−1 and un. Next we assign the labels to the vertices 4, 3, 2, 2, 2, 1 to
the vertices vn−2, vn−1, xn−2, xn−1, wn−2, and wn−1 respectively. Finally we assign the labels
4, 2, 3 respectively to the vertices yn−3, yn−2 and yn−1.
Case 4. n ≡ 3 (mod 4).
Let n = 4t + 3, t ∈ N. Assign the label to the vertices ui, xi, yi, wi (1 ≤ i ≤ n − 3) and
vi (1 ≤ i ≤ n − 4) as in Case 3. Now we assign the labels 4, 3, 3, 3, 1, 1, 3, 4, 3 respectively
to the vertices un−2, un−1, un, xn−2, xn−1, yn−2, yn−1, wn−2 and wn−2. Finally we assign the
labels 2, 4, 3 to the vertices vn−3, vn−2 and vn−1 respectively. 2Theorem 3.7 The subdivision of ladder S(Ln) is 4-total prime cordial.
Proof Let V (Ln) = {ui, vi : 1 ≤ i ≤ n} and E(Ln) = {uiui+1, vivi+1 : 1 ≤ i ≤ n − 1} ∪{uivi : 1 ≤ i ≤ n}. Let yi, wi and xi be the vertices which subdivide the edges uiui+1, uivi and
vivi+1 respectively. Clearly |V (S(Ln))| + |E(S(Ln))| = 11n− 6.
Case 1. n ≡ 0 (mod 4).
Let n = 4t, t ∈ N. Assign the label 4 to the vertices u1, u2, · · · , ut and v1, v2, · · · , vt.
98 R. Ponraj, J. Maruthamani and R. Kala
Assign the label 2 to the vertices ut+1, ut+2, · · · , u2t and vt+1, vt+2, · · · , v2t. Next we assign
the label 3 to the vertices u2t+1, u2t+2,. . . , u3t and v2t+1, v2t+2, · · · , v3t then we assign the
label 1 to the vertices u3t+1, u3t+2, · · · , un−1 and v3t+1, v3t+2, · · · , vn−1. Finally, we assign
the labels 4 and 3 to the vertices un and vn respectively. Next we consider the vertices xi
(1 ≤ i ≤ n). Assign the label 4 to the vertices x1, x2, · · · , xt and assign the label 2 to the
vertices xt+1, xt+2, · · · , x2t. Now we assign the label 3 to the vertices x2t+1, x2t+2, · · · , x3t.
Finally we assign the label 1 to the vertices x3t+1, x3t+2, · · · , xn. Now we consider the vertices
yi, wi (1 ≤ i ≤ n−1). Assign the label 4 to the vertices y1, y2, · · · , yt and w1, w2, · · · , wt. Assign
the label 2 to the vertices yt+1, yt+2, · · · , y2t−1 and wt+1, wt+2, · · · , w2t−1. Next we assign the
label 3 to the vertices y2t, y2t+1, · · · , y3t−1 and w2t, w2t+1, · · · , w3t−1 then we assign the label
1 to the vertices y3t, y3t+1, · · · , yn−2 and w3t, w3t+1,· · · , wn−2, wn−1. Finally, we assign the
labels 2 vertex yn−1.
Case 2. n ≡ 1 (mod 4).
Let n = 4t+ 1, t ∈ N. As in Case 1, assign the label to the vertices ui (1 ≤ i ≤ n− 1), vi
(1 ≤ i ≤ n − 1), xi (1 ≤ i ≤ n), yi (1 ≤ i ≤ n − 2) and wi (1 ≤ i ≤ n − 2). Finally we assign
the labels 2, 2, 4, 3 respectively to the vertices un, vn, yn−1 and wn−1.
Case 3. n ≡ 2 (mod 4).
Let n = 4t+ 2, t ∈ N. As in Case 2, assign the label to the vertices ui (1 ≤ i ≤ n− 1), vi
(1 ≤ i ≤ n− 1), xi (1 ≤ i ≤ n− 1), yi (1 ≤ i ≤ n− 2) and wi (1 ≤ i ≤ n− 2). Finally we assign
the labels 4, 3, 2, 4, 3 to the vertices un, vn, xn, yn−1 and wn−1 respectively.
Case 4. n ≡ 3 (mod 4).
Let n = 4t+ 3, t ∈ N. As in Case 3, assign the label to the vertices ui (1 ≤ i ≤ n− 1), vi
(1 ≤ i ≤ n− 1), xi (1 ≤ i ≤ n− 1), yi (1 ≤ i ≤ n− 2) and wi (1 ≤ i ≤ n− 2). Finally we assign
the labels 3, 4, 2, 3, 4 respectively to the vertices un, vn, xn, yn−1 and wn−1. 2Theorem 3.8 The subdivision of double triangular snake S(DTn) is 4-total prime cordial.
Proof Let Pn be the path u1u2 · · ·un. Let vi, wi be the vertex adjacent to uiui+1. Let xi,
yi, zi, si and ri be the vertex which subdivide the edges uiui+1, uivi, viui+1, uiwi and wiui+1
respectively. Clearly |V (S(DTn))| + |E(S(DTn))| = 18n− 17.
Case 1. n ≡ 0 (mod 4), n ≥ 8.
Let n = 4t, t ∈ N. Assign the label 4 to the vertices u1, u2, · · · , ut and assign the label 2 to
the vertices ut+1, ut+2, · · · , u2t. Next we assign the label 3 to the vertices u2t+1, u2t+2, · · · , u3t
then we assign the label 1 to the vertices u3t+1, u3t+2, · · · , un−1. Finally, we assign the label 3
to the vertex un. Now we consider the vertices vi,wi (1 ≤ i ≤ n− 1). Assign the label 4 to the
vertices v1, v2, · · · , vt and w1, w2, · · · , wt. Assign the label 2 to the vertices vt+1, vt+2, · · · , v2t−1
and wt+1, wt+2, · · · , w2t−1. Next we assign the label 3 to the vertices v2t, v2t+1, · · · , v3t−1
and w2t, w2t+1, · · · , w3t−1 then we assign the label 1 to the vertices v3t, y3t+1, · · · , vn−3 and
w3t, w3t+1, · · · , wn−3. Finally we assign the labels 2, 4, 2, 4 respectively to the vertices vn−2,
vn−1, wn−2 and wn−1. Next we move to the vertices xi (1 ≤ i ≤ n− 1). Assign the label 4 to
4-Total Prime Cordiality of Certain Subdivided Graphs 99
the vertices x1, x2, · · · , xt and assign the label 2 to the vertices xt+1, ut+2, · · · , x2t−1. Next we
assign the label 3 to the vertices x2t, x2t+1, · · · , x3t−1 then we assign the label 1 to the vertices
x3t, x3t+1, · · · , xn−2. Finally, we assign the label 3 to the vertex xn−1. Now we consider the
vertices yi,si (1 ≤ i ≤ n− 1). Assign the label 4 to the vertices y1, y2, · · · , yt and s1, s2, · · · , st.
Assign the label 2 to the vertices yt+1, yt+2, · · · , y2t and st+1, st+2, · · · , s2t. Next we assign the
label 3 to the vertices y2t+1, y2t+2, · · · , y3t−1 and s2t+1, s2t+2, · · · , s3t−1. Finally we assign the
label 1 to the vertices y3t, y3t+1, · · · , yn−1 and s3t, s3t+1, · · · , sn−1. Next we move to the vertices
zi, ri (1 ≤ i ≤ n − 1). Assign the label 4 to the vertices z1, z2, · · · , zt−1 and r1, r2, · · · , rt−1.
Assign the label 2 to the vertices zt, zt+1, · · · , z2t−1 and rt, rt+1, · · · , r2t−1. Next we assign the
label 3 to the vertices z2t, z2t+1, · · · , z3t−1 and r2t, r2t+1, · · · , r3t−1. Finally we assign the label
1 to the vertices z3t, z3t+1, · · · , zn−1 and r3t, r3t+1, · · · , rn−1. Clearly tf (1) = tf (2) = tf (3) =
18t− 4 and tf (4) = 18t− 5.
Case 2. n ≡ 1 (mod 4), n ≥ 9.
Let n = 4t+ 1, t ∈ N. As in Case 1, assign the label to the vertices ui (1 ≤ i ≤ n− 1), vi,
wi, xi, yi, zi, si, ri (1 ≤ i ≤ n− 2). Finally we assign the labels 4, 4, 2, 3, 1, 4, 3, 2 respectively
to the vertices un, vn−1, wn−1, xn−1, yn−1, zn−1, sn−1 and rn−1. Obviously tf (1) = 18t + 1
and tf (2) = tf (3) = tf (4) = 18t.
Case 3. n ≡ 2 (mod 4), n ≥ 10.
Let n = 4t+ 2, t ∈ N. Assign the label to the vertices ui (1 ≤ i ≤ n− 2), vi, wi, xi, yi, zi
(1 ≤ i ≤ n− 3), si, ri (1 ≤ i ≤ n− 2) by in case 1. Finally we assign the labels 2, 4, 3, 4, 3, 3,
1, 2, 3, 4, 2, 4, 1, 4 to the vertices un−1, un, vn−2, vn−1, wn−2, wn−1, xn−2, xn−1, yn−2, yn−1,
zn−2, zn−1, sn−1 and rn−1 respectively. It is easy to verify that tf (1) = tf (2) = tf (3) = 18t+5
and tf (4) = 18t+ 4.
Case 4. n ≡ 3 (mod 4), n ≥ 11.
Let n = 4t + 3, t ∈ N. As in Case 3, assign the label to the vertices ui (1 ≤ i ≤ n − 1),
vi, wi, xi, yi, zi, si, ri (1 ≤ i ≤ n − 2). Finally we assign the labels 3, 3, 2, 2, 4, 3, 4,
1 respectively to the vertices un, vn−1, wn−1, xn−1, yn−1, zn−1, sn−1 and rn−1. Clearly
tf (1) = tf (2) = tf (4) = 18t+ 9 and tf (3) = 18t+ 10.
Case 5. t = 2, 3, 4, 5, 6, 7.
A 4-total prime cordial labeling is given in Table 1.
n 2 3 4 5 6 7
u1 3 4 4 4 4 4
u2 4 2 2 4 4 4
u3 3 3 2 2 2
u4 3 3 3 2
u5 1 1 3
100 R. Ponraj, J. Maruthamani and R. Kala
u6 2 1
u7 1
v1 4 4 4 4 4 4
v2 3 3 2 2 4
v3 4 3 3 2
v4 3 3 3
v5 4 3
v6 1
w1 2 4 4 4 4 4
w2 1 2 2 2 4
w3 1 3 3 2
w4 1 1 3
w5 3 3
w6 1
x1 1 2 4 4 4 4
x2 3 2 2 2 2
x3 3 3 3 2
x4 3 3 3
x5 4 3
x6 1
y1 3 4 4 4 4 4
y2 3 3 2 2 4
y3 1 3 3 2
y4 1 1 3
y5 3 3
y6 1
z1 4 2 2 4 4 4
z2 3 3 2 2 2
z3 4 3 3 2
z4 1 1 3
z5 4 1
z6 1
s1 3 4 4 4 4 4
s2 1 2 2 2 4
4-Total Prime Cordiality of Certain Subdivided Graphs 101
s3 1 3 3 2
s4 1 1 3
s5 3 3
s6 1
r1 2 2 2 4 4 4
r2 1 3 2 2 2
r3 1 3 3 2
r4 1 1 3
r5 2 3
r6 1
Table 1
This completes the proof. 2References
[1] I.Cahit, Cordial graphs:A weaker version of graceful and harmonious graphs, Ars Combi-
natoria, 23(1987), 201-207.
[2] J.A.Gallian, A Dynamic survey of graph labeling, The Electronic Journal of Combinatorics,
19 (2017) #Ds6.
[3] F.Harary, Graph Theory, Addision wesley, New Delhi (1969).
[4] R.Ponraj, J.Maruthamani and R.Kala, k-Total prime cordial labeling of graphs, Journal
of Algorithms and Computation, 50(1), 143-149.
[5] R.Ponraj, J.Maruthamani and R.Kala, Some 4-total prime cordial labeling of graphs, Jour-
nal of Applied Mathematics and Informatics, 37(1-2)(2019), 149-156.
[6] R.Ponraj, J.Maruthamani and R.Kala, Some classes of 4-total prime cordial labeling of
graphs, Global Journal of Engineering science and Researches, 5(11), 319-327.
[7] R.Ponraj, J.Maruthamani and R.Kala, 4-total prime cordial labeling of some special graphs,
Jordan Journal of Mathematics and Statistics, (Accepted for Publication).
[8] R.Ponraj, J.Maruthamani and R.Kala, 4-Total prime cordial labeling of some cycle related
graphs, Journal of Algorithms and Computation, 50(2), 49-57.
[9] R.Ponraj and J.Maruthamani, 4-total prime cordial labeling of some derived graphs, Ars
Combinatoria, (Accepted for Publication).
International J.Math. Combin. Vol.4(2019), 102-111
The Signed Product Cordial for Corona of Paths and
Fourth Power of Paths
S. Nada, A. Elrokh
(Department of Mathematics, Faculty of Science, Menoyfia University, Egypt)
A. Elrayes and A. Rabie
(Institute of National Planning, Cairo, Egypt)
E-mail: [email protected] [email protected], [email protected] and [email protected]
Abstract: A graph G = (V, E) is called signed product cordial if it is possible to label the
vertex by the function f : V → {−1, 1} and label the edges by f∗ : E → {−1, 1}, where
f∗(uv) = f(u) · f(v), u, v ∈ V so that |v−1 − v1| 6 1 and |e−1 − e1| 6 1. In [3] J.Devaraj and
P.Delphy, they have defined signed graphs, and they have started by labeling edges and then
induced the labeling of vertices. In this paper, we contribute some new results on signed
product cordial labeling and present necessary and sufficient conditions for signed product
cordial for corona of paths and fourth power of paths.
Key Words: Second power, fourth power, corona graph, signed product cordial graph,
Smarandachely signed product cordial labeling.
AMS(2010): 05C78,05C75,05C20.
§1. Introduction
The labeling of graphs is perceived to be a primarily theoretical subject in the field of graph the-
ory and discrete mathematics, it serves as models in a wide range of application like astronomy,
coding theory, circuit design and communication networks addressing. The concept of graph
labeling was introduced during the sixties’ of the last century by Rosa [12]. Many researches
have been working with different types of labeling graphs [1,4,5]. In 1954 Harray introduced
S-cordiality [10]. An excellent reference for this purpose is the survey written by Gallian [6].
All graphs considered in this theme are finite, simple and undirected. The original concept of
cordial graphs is due to Chait[2]. He showed that each tree is cordial; an Euerlian graph is not
cordial if its size is congruent to 2(mod 4). Let G = (V,E) be a graph and let f : V → {−1, 1}be a labeling of its vertices, and let the induced edge labeling f∗ : E → {−1, 1} be given by
f∗(e) = (f(u) · f(v)), where e = uv and u, v ∈ V .
Let v−1 and v1 be the numbers of vertices that are labeled by −1 and 1, respectively, and
let e−1 and e1 be the corresponding numbers of edges. Such a labeling is called signed product
1Received April 14, 2019, Accepted December 6, 2019.
The Signed Product Cordial for Corona of Paths and Fourth Power of Paths 103
cordial if both |v−1 − v1| 6 1 and |e−1 − e1| 6 1 hold. Otherwise, it is called Smarandachely
signed product cordial if |v−1 − v1| > 1 or |e−1 − e1| > 1. The corona G1
⊙G2 of two graphs
G1 (with n1 vertices, m1 edges) and G2 (with n2 vertices, m2 edges) is defined as the graph
obtained by taking one copy of G1 and copies of G2, and then joining the ith vertex of G1 with
an edge to every vertex in the ith copy of G2. It is easy to see that the corona G1
⊙G2 that
has n1 + n1n2 vertices and m1 + n1m2 + n1n2 edges. The fourth power of a paths Pn, denoted
by P 4n , is Pn
⋃J , where J is the set of all edges of the form edges vivj such that 2 ≤ d(vivj) ≤ 4
and i < j where d(vivj) is the shortest path from vi to vj .
§2. Terminologies and Notations
A path with m vertices and m−1 edges, denoted by Pm, and its fourth power P 4n has n vertices
and 4n − 10 edges. We let L4r denote the labeling (−1)211 (−1)211· · · (−1)211 (repeated r-
times), Let L′4r denote the labeling (−1)11(−1) (−1)11(−1)· · · (−1)11(−1) (repeated r-times).
The labeling 11(−1)2 11(−1)2· · · 11(−1)2 (repeated r-times) and labeling 1(−1)211(−1)21
· · · 1(−1)21 (repeated r-times) are written S4r and S′4r. Let Mr denote the labeling (−1)1
(−1)1 · · · (−1)1, zero-one repeated rtimes if r is even and (−1)1 (−1)1 · · · (−1)1(−1) if r is
odd; for example, M6 = (−1)1(−1)1(−1)1 and M5 = (−1)1(−1)1(−1). We let M ′r denote the
labeling 1(−1)1(−1) · · · 1(−1). Sometimes, we modify the labeling Mr or M ′r by adding symbols
at one end or the other (or both). Also, L4r (or L′4r ) with extra labeling from right or left (or
both sides). If L is a labeling for a path pm and M is a labeling for fourth power of path Pn,
then we use the notation [L;M ] to represent the labeling of the corona Pm
⊙P 4
n . Additional
notation that we use is the following: for a given labeling of the corona Pm
⊙P 4
n , we let vi
and ei (for i = −1, 1) be the numbers of vertices and edges, respectively, that are labeled by i
of the corona Pm
⊙P 4
n , and let xi and ai be the corresponding quantities for pm, and we let
yi and bi be those for P 4n , which are connected with vertices labeled (−1) of Pm. Similarly, let
y′i and b′i for P 4n which are connected with vertices labeled 1 of Pm. It is easy to verify that
v−1 = x−1+x−1y−1+x1y′−1, v1 = x1+x−1y1+x1y
′1, e−1 = a−1+x−1b−1+x1b
′−1+x−1y−1+x1y
′1
and e1 = a1+x−1b1+x1b′1+x−1(x−1y1)+x1y
′−1. Thus, v−1−v1 = (x−1−x1)+x−1(y−1−y1)+
x1(y′−1−y′1) and e−1−e1 = (a−1−a1)+x−1(b−1−b1)+x1(b
′−1−b′1)+x−1(y−1−y1)−x1(y
′−1−y′1).
When it comes to the proof, we only need to show that, for each specified combination of
labeling, |v−1 − v1| ≤ 1 and |e−1 − e1| ≤ 1.
§3. Main Results
In this section, we show that the corona Pm
⊙P 4
n is signed product cordial for all m, n ≥ 7 and
also for n = 3 with m ≥ 1. This target will be achieved after the following series of lemmas.
Lemma 3.1 The corona Pm
⊙P 4
3 is signed product cordial if and only if m 6= 1.
Proof Suppose that n = 3. The following cases will be examined.
104 S. Nada, A. Elrokh, A. Elrayes and A. Rabie
Case 1. Obviously, P1
⊙P 4
3 isomorphic to the complete graph K4. Since K4 is not cordial,
P1
⊙P 4
3 is not is signed product cordial.
Case 2. Suppose that m = 2. Then we label the vertices of P2
⊙P 4
3 by [−11;−11−1, 1−11];
hence v−1 − v1 = 0 and e−1 − e1 = 1. So, P2
⊙P 4
3 is signed product cordial.
Case 3. Suppose that m = 3. Then we label the vertices of P3
⊙P 4
3 by [−1 − 1 −1;−111, 111,−11 − 1]; hence v−1 − v1 = 0 and e−1 − e1 = 0. So, P3
⊙P 4
3 is signed prod-
uct cordial.
Case 4. m ≡ 0(mod4).
Suppose that m = 4r, r ≥ 1. We choose the labeling [L4r;−11 − 1,−11 − 1, 1 − 11, 1 −11, · · · , (r − times)] for P4r
⊙P 4
3 . Therefore x−1 = x1 = 2r, a−1 = 2r − 1, a1 = 2r, y−1 = 2,
y1 = 1, y′−1 = 1, y′1 = 2, b−1 = 2, b′−1 = 2, b1 = 1 and b′1 = 1. Hence v−1 − v1 =
(x−1 − x1) + x−1.(y−1 − y1) + x1.(y′−1 − y′1) = 0 and e−1 − e1 = (a−1 − a1) + x−1.(b−1 − b1) +
x1.(b′−1 − b′1) + x−1.(y−1 − y1)− x1.(y
′−1 − y′1) = −1. Thus P4r
⊙P 4
3 is signed product cordial.
Case 5. m ≡ 1(mod4).
Suppose that m = 4r+1, r ≥ 1. We choose the labeling [L4r1;−11−1, −11−1, 1−11, 1−11, · · · , (r − times),−11 − 1] for P4r+1
⊙P 4
3 . Therefore x−1 = 2r, x1 = 2r + 1, a−1 = 2r − 1,
a1 = 2r + 1, and for the first 4r- vertices y−1 = 2, y1 = 1, y′−1 = 1, y′1 = 2, b−1 = b′−1 = 2
and b1 = b′1 = 1, and for the cycle c3 which is connected to last vertex in P4r+1, we have
y′′−1 = 2, y′′1 = 1, b′′−1 = 2 and b′′(1) = 1, where y′′i and b′′i are the numbers of vertices and
edges labeled by i in P 43 that is connected to the last vertex of P4r+1. It is easily to verify that
v−1 = x−1 + x−1.y−1 + (x1 − 1).y′−1 + y′′−1 = 8r + 2, v1 = x1 + x−1.y1 + (x1 − 1).y′1 + y′′1 =
8r + 2, e−1 = a−1 + x−1.b−1 + (x1 − 1).b′−1 + x−1.y−1 + (x1 − 1).y′1 + b′′−1 + 1 = 14r + 3 and
e1 = a1 + x−1.b1 + (x1 − 1).b′1 + x−1.y1 + (x1 − 1).y′−1 + b′′(1) + 2 = 14r + 3. It follows that
v−1 − v1 = 0 and e−1 − e1 = 0. Thus P4r+1
⊙P 4
3 is cordial.
Case 6. m ≡ 2(mod4).
Suppose that m = 4r+2, r ≥ 1. We choose the labeling [L4r1(−1); (−1)1(−1), (−1)1(−1),
1(−1)1, 1(−1)1, ..., (r − times), 1(−1)1, (−1)1(−1)] for P4r+2
⊙P 4
3 . Therefore x−1 = x1 =
2r + 1, a−1 = 2r, a1 = 2r + 1, y−1 = 2, y1 = 1, y′−1 = 1, y′1 = 2, b−1 = 2, b′−1 = 2,
b1 = 1 and b′1 = 1. Hence v−1 − v1 = (x−1 − x1) + x−1.(y−1 − y1) + x1.(y′−1 − y′1) = 0 and
e−1 − e1 = (a−1 − a1) + x−1.(b−1 − b1) + x1.(b′−1 − b′1) + x−1.(y−1 − y1) − x1.(y
′−1 − y′1) = −1.
Thus P4r+2
⊙P 4
3 is cordial.
Case 7. m ≡ 3(mod4).
Let m = 4r + 3, r ≥ 1. We choose the labeling [L4r1(−1)(−1); (−1)1(−1), (−1)1(−1),
1(−1)1, 1(−1)1, · · · , (r − times), 1(−1)1, (−1)1(−1), 1(−1)1] for P4r+3
⊙P 4
3 . Therefore x−1 =
2r+2, x1 = 2r+1, a−1 = 2r, a1 = 2r+2, and for the first 4r- vertices y−1 = 2, y1 = 1, y′−1 = 1,
y′1 = 2, b−1 = b′−1 = 2 and b1 = b′1 = 1, and for the cycle P 43 which is connected to last vertex of
P4r+3, we have y′′0 = 1, y′′1 = 2, b′′0 = 2 and b′′1 = 1, where y′′i and b′′i are the numbers of vertices
and edges labeled by i in P 43 that is connected to the last vertex of P4r+3. Similar to Case 2, we
The Signed Product Cordial for Corona of Paths and Fourth Power of Paths 105
conclud that v−1 − v1 = (x−1 −x1)+x−1.(y−1 − y1)+ (x1 − 1).(y′−1 − y′1)+ (y′′−1 − y′′1 ) = 0 and
e−1 − e1 = (a−1 − a1) + x−1.(b−1 − b1) + (x1 − 1).(b′−1 − b′1) + x−1.(y−1 − y1)− (x1 − 1).(y′−1 −y′1) + (c−1 − c1) − 1 = 0. Hence P4r+3
⊙P 4
3 is cordial. Thus the lemma is proved. 2Lemma 3.2 If n ≡ 0(mod4), then Pm
⊙P 4
n is signed product cordial for all m ≥ 1.
Proof Suppose that n = 4s, where s ≥ 2. The following cases will be examined.
Case 1. Suppose that m = 1. Then we label the vertices of P1
⊙P 4
4s by [−1;−1L4s−4 − 112].
Therefore x−1 = 1, x1 = 0, a−1 = a1 = 0, y−1 = y1 = 2s, b−1 = b1 = 8s − 5. It follows that
v−1 − v1 = 1 and e−1 − e1 = 0. As an example, Figure 1 illustrates P1
⊙P 4
8 . Hence, P1
⊙P 4
4s
is signed product cordial.
-1 -1 -1 1 1 -1 1 1
-1
Figure 1
Case 2. Suppose that m = 2. Then we label the vertices of P2
⊙P 4
4s by [−11;−1L4s−4 −112, 12L
′4s−4−12]. Therefore x−1 = x1 = 1, a−1 = 1, a1 = 0, y−1 = y1 = 2s, b−1 = b1 =
8s− 5, y′−1 = y′1 = 2s and b′−1 = b′1 = 8s− 5. It follows that v−1 − v1 = 0 and e−1 − e1 = 1. As
an example, Figure 2 illustrates P2
⊙P 4
8 . Hence, P2
⊙P 4
4s is signed product cordial.
-1 -1 -1 1 1 1 1 1 1 1 -1 1 1 -1 -1 -1
-1 1
Figure 2
106 S. Nada, A. Elrokh, A. Elrayes and A. Rabie
Case 3. Suppose that m = 3. Then we label the vertices of P3
⊙P 4
4s by [−121;−1L4s−4 −112,−1L4s−4 − 112, 12L
′4s−4−12]. Therefore x−1 = 2, x1 = 2, a−1 = a1 = 1, y−1 = y1 =
2s, b−1 = b1 = 8s− 5, y′−1 = y′1 = 2s and b′−1 = b′1 = 8s− 5. It follows that v0−1 − v1 = 1 and
e−1 − e1 = 0. As an example, Figure 3 illustrates P3
⊙P 4
8 . Hence, P3
⊙P 4
4s is signed product
cordial.
-1 -1 -1 1 1 -1 1 1 1 1 -1 1 1 -1 -1 -1
-1 -1 1
-1 -1 -1 1 1 -1 1 1
Figure 3
Case 4. m = 0(mod 4).
Suppose thatm = 4r, where r ≥ 2. Then we label the vertices of P4r
⊙P 4
4s by [L4r;−1L4s−4−112,−1L4s−4 − 112, 12L
′4s−4−12,
12L′4s−4−12, ..., (r − time)]. Therefore x−1 = x1 = 2r, a−1 = 2r − 1, a1 = 2r, y−1 = y1 =
2s, b−1 = b1 = 8s− 5, y′−1 = y′1 = 2s and b′−1 = b′1 = 8s− 5. It follows that v−1 − v1 = 0 and
e−1−e1 = 1. As an example, Figure 4 illustrates P4
⊙P 4
8 . Hence, P4r
⊙P 4
4s is signed product
cordial.
-1 -1 -1 1 1 -1 1 1 1 1 -1 1 1 -1 -1 -1 1 1 -1 1 1 -1 -1 -1
-1 -1 1 1
-1 -1 -1 1 1 -1 1 1
Figure 4
Case 5. m = 1(mod 4).
Suppose that m = 4r + 1, where r ≥ 1. Then we label the vertices of P4r+1
⊙P 4
4s by
[L4r−1;−1L4s−4−112,−1L4s−4−112, 12L′4s−4−12, 12L
′4s−4−12, · · · , (r−time),−1L4s−4−112].
Therefore x−1 = 2r+1, x1 = 2r, a−1 = a1 = 2r, y−1 = y1 = 2s, b−1 = b1 = 8s−5, y′−1 = y′1 = 2s
The Signed Product Cordial for Corona of Paths and Fourth Power of Paths 107
and b′−1 = b′1 = 8s− 5. It follows that v−1 − v1 = 1 and e−1 − e1 = 0. Hence, P4r+1
⊙P 4
4s is
signed product cordial.
Case 6. m = 2(mod 4).
Suppose that m = 4r + 2, where r ≥ 1. Then we label the vertices of P4r+2
⊙P 4
4s by
[L4r−11;−1L4s−4−112,−1L4s−4−112, 12L′4s−4−12, 12L
′4s−4−12, · · · , (r−time), −1L4s−4−112,
12L′4s−4−12]. Therefore x−1 = x1 = 2r + 1, a−1 = 2r + 1, a1 = 2r, y−1 = y1 = 2s, b−1 = b1 =
8s − 5, y′−1 = y′1 = 2s and b′−1 = b′1 = 8s − 5. It follows that v−1 − v1 = 0 and e−1 − e1 = 1.
Hence, P4r+2
⊙P 4
4s is signed product cordial.
Case 7 m = 3(mod 4).
Suppose that m = 4r + 3, where r ≥ 1. Then we label the vertices of P4r+3
⊙P 4
4s by
[L4r−121;−1L4s−4−112,−1L4s−4−112, 12L′4s−4−12,
12L′4s−4v2, · · · , (r − time), −1L4s−4−112,−1L4s−4−112, 12L
′4s−4−12]. Therefore x−1 = 2r +
2, x1 = 2r + 1, a−1 = a1 = 2r + 1, y−1 = y1 = 2s, b−1 = b1 = 8s − 5, y′−1 = y′1 = 2s and
b′−1 = b′1 = 8s− 5. It follows that v−1 − v1 = 1 and e−1 − e1 = 0. Hence, P4r+3
⊙P 4
4s is signed
product cordial. 2Lemma 3.3 If n ≡ 1(mod4), then Pm
⊙P 4
n is cordial for all m ≥ 1.
Proof Suppose that n = 4s+ 1, where s ≥ 2. The following cases will be examined.
Case 1. Suppose thatm = 1. Then we label the vertices of P1
⊙P 4
4s+1 by [−1; 12L′4s−4−11−1].
Therefore x−1 = 1, x1 = 0, a−1 = a1 = 0, y−1 = 2s, y1 = 2s+ 1, b−1 = b1 = 8s− 3 . It follows
that v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P1
⊙P 4
4s+1 is signed product cordial.
Case 2. Suppose thatm = 2. Then we label the vertices of P2
⊙P 4
4s+1 by [−11;−12L4s−41−11,
12L′4s−4−11−1]. Therefore x−1 = x1 = 1, a−1 = 1, a1 = 0, y−1 = 2s + 1, y1 = 2s, b−1 = b1 =
8s − 3, y′−1 = 2s + 1, y′1 = 2s and b′−1 = b′1 = 8s − 3. It follows that v−1 − v1 = 0 and
e−1 − e1 = −1. Hence, P2
⊙P 4
4s+1 is signed product cordial.
Case 3. Let m = 3. Then we label the vertices of P3
⊙P 4
4s+1 by [−11−1;−12L4s−41−11,
12L′4s−4−11−1, 12L
′4s−4−11−1]. Therefore x−1 = 2, x1 = 1, a−1 = 2, a1 = 0, y−1 = 2s+1, y1 =
2s, b−1 = b1 = 8s − 3, y′−1 = 2s, y′1 = 2s + 1, b′−1 = b′1 = 8s − 3, y′′−1 = 2s, y′′1 = 2s + 1
andb′′−1 = b′′1 = 8s − 3. It follows that v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P3
⊙P 4
4s+1 is
signed product cordial.
Case 4. m = 0(mod 4).
Suppose that m = 4r, where r ≥ 1. Then we label the vertices of P4r
⊙P 4
4s+1 by
[M4r;−12L4s−41−11, 12L′4s−4−11−1,−12L4s−41−11,
12L′4s−4−11−1, · · · , (r − time)]. Therefore x−1 = x1 = 2r, a−1 = 4r − 1, a1 = 0, y−1 =
2s + 1, y1 = 2s, b−1 = b1 = 8s − 3, y′−1 = 2s, y′1 = 2s + 1 and b′−1 = b1 = 8s − 3. It fol-
lows that v−1 − v1 = 0 and e0 − e1 = −1. Hence, P4r
⊙P 4
4s+1 is signed product cordial.
Case 5. m = 1(mod 4).
Suppose that m = 4r + 1, where r ≥ 1. Then we label the vertices of P4r+1
⊙P 4
4s+1 by
108 S. Nada, A. Elrokh, A. Elrayes and A. Rabie
[M4r−1;−12L4s−41−11, 12L′4s−4−11−1,−12L4s−41−11, 12L
′4s−4−11−1, · · · , (r−time), 12L
′4s−4
−11−1]. Therefore x−1 = 2r + 1, x1 = 2r, a−1 = 4r, a1 = 0, y−1 = 2s + 1, y1 = 2s, b−1 = b1 =
8s− 3, y′−1 = 2s, y′1 = 2s+ 1, b′−1 = b′1 = 8s− 3, y′′−1 = 2s, y′′1 = 2s+ 1and b′′−1 = b′′1 = 8s− 3. It
follows that v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P4r+1
⊙P 4
4s+1 is signed product cordial.
Case 6. m = 2(mod 4).
Letm = 4r+2, where r ≥ 1. Label the vertices of P4r+2
⊙P 4
4s+1 by [M4r+2;−12L4s−41−11,
12L′4s−4−11−1,−12L4s−41−11, 12L
′4s−4−11−1,−12L4s−41−11, 12L
′4s−4−11−1, · · · , (r−time)].
Therefore x−1 = x1 = 2r+1, a−1 = 4r+1, a1 = 0, y−1 = 2s+1, y1 = 2s, b−1 = b1 = 8s−3, y′−1 =
2s, y′1 = 2s+ 1 and b′−1 = b′1 = 8s− 3. It follows that v−1 − v1 = 0 and e0 − e1 = −1. Hence
P4r+2
⊙P 4
4s+1 is signed product cordial.
Case 7. m = 3(mod 4).
Suppose that m = 4r + 3, where r ≥ 1. We then label the vertices of P4r+3
⊙P 4
4s+1
by [M4r+2−1; −12L4s−41−11, 12L′4s−4−11−1, −12L4s−41−11, 12L
′4s−4−11−1, −12L4s−41−11,
12L′4s−4−11−1, · · · , (r − time), 12L
′4s−4−11−1]. Therefore x−1 = 2r + 2, x1 = 2r + 1, a−1 =
4r + 1, a1 = 0, y−1 = 2s + 1, y1 = 2s, b−1 = b1 = 8s − 3, y′0 = 2s, y′1 = 2s + 1, b′0 = b′1 =
8s − 3, y′′−1 = 2s, y′′1 = 2s + 1 and b′′−1 = b′′1 = 8s − 3. It follows that v−1 − v1 = 0 and
e−1 − e1 = 1. Hence, P4r+3
⊙P 4
4s+1 is signed- cordial. 2Lemma 3.4 If n ≡ 2(mod4), then Pm
⊙P 4
n is cordial for all m ≥ 1.
Proof Suppose that n = 4s+ 2, where s ≥ 2. The following cases will be examined.
Case 1. Suppose thatm = 1. Then we label the vertices of P1
⊙P 4
4s+2 by [−1;−113−1s4s−4−1].
Therefore x−1 = 1, x1 = 0, a−1 = a1 = 0, y−1 = y1 = 2s+ 1, b−1 = b1 = 8s− 1. It follows that
v−1 − v1 = 1 and e−1 − e1 = 0. Hence, P1
⊙P 4
4s+2 is signed product cordial.
Case 2. Suppose that m = 2. Then, label the vertices of P2
⊙P 4
4s+2 by [−11;−113−1s4s−4−1,
−1l4s−4−113−1]. Therefore x−1 = x1 = 1, a−1 = 1, a1 = 0, y−1 = y1 = 2s + 1, b−1 = b1 =
8s−1, y′−1 = y′1 = 2s+1 and b′−1 = b′1 = 8s−1. It follows that v−1−v1 = 0 and e−1−e1 = −1.
Hence, P2
⊙P 4
4s+2 is signed product cordial.
Case 3. Let m = 3. Then we label the vertices of P3
⊙P 4
4s+2 by [−1−11; −113−1s4s−4−1,
−113−1s4s−4−1, −1l4s−4−113−1]. Therefore x−1 = 2, x1 = 1, a−1 = a1 = 1, y−1 = y1 =
2s+ 1, b−1 = b1 = 8s− 1, y′−1 = y′1 = 2s+ 1 and b′−1 = b′1 = 8s− 1. It follows that v−1 − v1 = 1
and e−1 − e1 = 0. Hence, P3
⊙P 4
4s+2 is signed product cordial.
Case 4. m = 0(mod 4).
Suppose that m = 4r, where r ≥ 1. Then we label the vertices of P4r
⊙P 4
4s+2 by
[l4r;−113−1s4s−4−1,−113−1s4s−4−1,−1L4s−4−113−1,−1L4s−4−113−1, · · · , (r−time)]. There-
fore x−1 = x1 = 2r, a−1 = 2r−1, a1 = 2r, y−1 = y1 = 2s+1, b−1 = b1 = 8s−1, y′−1 = y′1 = 2s+1
and b′−1 = b1 = 8s− 1. It follows that v−1 − v1 = 0 and e−1 − e1 = −1. Hence, P4r
⊙P 4
4s+2 is
signed product cordial.
Case 5. m = 1(mod 4).
The Signed Product Cordial for Corona of Paths and Fourth Power of Paths 109
Suppose that m = 4r + 1, where r ≥ 1. Then we label the vertices of P4r+1
⊙P 4
4s+2 by
[l4r−1;−113−1s4s−4−1,−113−1s4s−4−1,
−1L4s−4−113−1,−1L4s−4−113−1, · · · , (r − time), −113−1s4s−4−1]. Therefore x−1 = 2r +
1, x1 = 2r, a−1 = a1 = 2r, y−1 = y1 = 2s + 1, b−1 = b1 = 8s − 1, y′−1 = y′1 = 2s + 1 and
b′−1 = b′1 = 8s − 1. It follows that v−1 − v1 = 1 and e−1 − e1 = 0. Hence, P4r+1
⊙P 4
4s+2 is
signed- cordial.
Case 6. m = 2(mod 4).
Suppose that m = 4r + 2, where r ≥ 1. Then we label the vertices of P4r+2
⊙P 4
4s+2 by
[l4r−11;−113−1s4s−4−1,−113−1s4s−4−1,
−1l4s−4−113−1,−1l4s−4−113−1, · · · , (r − time), −113−1s4s−4−1,−1L4s−4−113−1]. There-
fore x−1 = x1 = 2r + 1, a−1 = 2r + 1, a1 = 2r, y−1 = y1 = 2s + 1, b−1 = b1 = 8s − 1, y′−1 =
y′1 = 2s + 1 and b′−1 = b′1 = 8s − 1. It follows that v−1 − v1 = 0 and e−1 − e1 = 1. Hence,
P4r+2
⊙P 4
4s+2 is signed product cordial.
Case 7. m = 3(mod 4).
Suppose that m = 4r + 3, where r ≥ 1. Then we label the vertices of P4r+3
⊙P 4
4s+2
by [L4r−121;−113−1s4s−4−1, −113−1s4s−4−1, −1L4s−4−113−1, −1L4s−4−113−1, · · · , (r −time), −113−1s4s−4−1, −113−1s4s−4−1,−1L4s−4−113−1]. Therefore x−1 = 2r + 2, x1 =
2r + 1, a−1 = a1 = 2r + 1, y−1 = y1 = 2s + 1, b−1 = b1 = 8s − 1, y′−1 = y′1 = 2s + 1 and
b′−1 = b′1 = 8s − 1. It follows that v−1 − v1 = 1 and e−1 − e1 = 0. Hence, P4r+3
⊙P 4
4s+2 is
signed product cordial. 2Lemma 3.5 If n ≡ 3(mod4), then Pm
⊙P 4
n is signed product cordial for all m ≥ 1.
Proof Suppose that n = 4s+ 3, where s ≥ 2. The following cases will be examined.
Case 1. Suppose that m = 1. Then we label the vertices of P1
⊙P 4
4s+3 by [−1; 12s4s−1].
Therefore x−1 = 1, x1 = 0, a−1 = a1 = 0, y−1 = 2s + 1, y1 = 2s + 2, b−1 = b1 = 8s + 1 . It
follows that v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P1
⊙P 4
4s+3 is signed product cordial.
Case 2. Suppose thatm = 2. Then, label the vertices of P2
⊙P 4
4s+3 by [−11;−121L4s, 12s4s−1].
Therefore x−1 = x1 = 1, a−1 = 1, a1 = 0, y−1 = 2s + 2, y1 = 2s+ 1, b−1 = b1 = 8s + 1, y′−1 =
2s+1, y′1 = 2s+2 and b′−1 = b′1 = 8s+1. It follows that v−1− v1 = 0 and e−1− e1 = 1. Hence,
P2
⊙P 4
4s+3 is signed product cordial.
Case 3. Suppose that m = 3. Then we label the vertices of P3
⊙P 4
4s+3 by [−11−1; −121L4s,
12s4s−1, 12s4s−1]. Therefore x−1 = 2, x1 = 1, a−1 = 2, a1 = 0, y−1 = 2s+ 2, y1 = 2s+1, b−1 =
b1 = 8s + 1, y′−1 = 2s + 1, y′1 = 2s + 2, b′−1 = b′1 = 8s + 1, y′′−1 = 2s + 1, y′′1 = 2s + 2
andb′′−1 = b′′1 = 8s + 1. It follows that v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P3
⊙P 4
4s+3 is
signed product cordial.
Case 4. m = 0(mod 4).
Suppose that m = 4r, where r ≥ 1. Then we label the vertices of P4r
⊙P 4
4s+3 by
[M4r;−121L4s, 12s4s−1,−121L4s, 12s4s−1, · · · , (r − time)]. Therefore x−1 = x1 = 2r, a−1 =
110 S. Nada, A. Elrokh, A. Elrayes and A. Rabie
4r − 1, a1 = 4r, y−1 = 2s + 2, y1 = 2s + 1, b−1 = b1 = 8s + 1, y′−1 = 2s + 1, y′1 = 2s + 2 and
b′−1 = b′1 = 8s + 1. It follows that v−1 − v1 = 0 and e−1 − e1 = −1. Hence, P4r
⊙P 4
4s+3 is
signed product cordial.
Case 5. m = 1(mod 4).
Suppose that m = 4r + 1, where r ≥ 1. Then we label the vertices of P4r+1
⊙P 4
4s+3
by [M4r−1;−121L4s, 12s4s−1,−121L4s, 12s4s−1, · · · , (r − time), 12s4s−1]. Therefore x−1 =
2r+1, x1 = 2r, a−1 = 4r, a1 = 0, y−1 = 2s+2, y1 = 2s+1, b−1 = b1 = 8s+1, y′−1 = 2s+1, y′1 =
2s + 2, b′−1 = b′1 = 8s + 1, y′′−1 = 2s + 1, y′′1 = 2s + 2and b′′−1 = b′′1 = 8s + 1. It follows that
v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P4r+1
⊙P 4
4s+3 is signed product cordial.
Case 6. m = 2(mod 4).
Suppose that m = 4r + 2, where r ≥ 1. Then we label the vertices of P4r+2
⊙P 4
4s+3
by [M4r+2;−121L4s, 12s4s−1,−121L4s, 12s4s−1,−121L4s, 12s4s−1, · · · , (r − time)]. Therefore
x−1 = x1 = 2r + 1, a−1 = 4r + 1, a1 = 0, y−1 = 2s+ 2, y1 = 2s + 1, b−1 = b1 = 8s + 1, y′−1 =
2s + 1, y′1 = 2s + 2 and b′−1 = b′1 = 8s + 1. It follows that v−1 − v1 = 0 and e−1 − e1 = −1.
Hence, P4r+2
⊙P 4
4s+3 is signed product cordial.
Case 7. m = 3(mod 4).
Suppose that m = 4r + 3, where r ≥ 1. Then we label the vertices of P4r+3
⊙P 4
4s+3 by
[M4r+2−1;−121L4s, 12s4s−1,−121L4s, 12s4s−1,
−121L4s, 12s4s−1..., (r − time)
12s4s−1]. Therefore x−1 = 2r + 2, x1 = 2r + 1, a−1 = 4r + 2, a1 = 0, y−1 = 2s + 2, y1 =
2s+ 1, b−1 = b1 = 8s+ 1, y′−1 = 2s+ 1, y′1 = 2s+ 2, b′−1 = b′1 = 8s+ 1, y′′−1 = 2s+ 1, y′′1 = 2s+ 2
and b′′−1 = b′′1 = 8s+ 1. It follows that v−1 − v1 = 0 and e−1 − e1 = 1. Hence, P4r+3
⊙P 4
4s+3
is signed product cordial. 2As a consequence of all lemmas mentioned above we conclude that the signed product
cordial of the corona between paths and fourth power of paths is cordial for all m, n ≥ 7 and
n = 3 for all m ≥ 1.
Theorem 3.1 The corona between paths and fourth power of paths is signed product cordial
for all m, n ≥ 7 and also for n = 3 with m ≥ 1.
Proof The proof follows directly from Lemma 3.1 to Lemma 3.5. 2Acknowledgement. The authors are thankful to the anonymous referee for useful suggestions
and valuable comments.
References
[1] I. Cahit, Cordial graphs – A weaker version of graceful and harmonious graphs, Ars Com-
binatoria, 23(1987), 201-207.
[2] I. Cahit, On cordial and 3-eqtitable labeling of graphs, Utiliuies Math., 37(1990).
The Signed Product Cordial for Corona of Paths and Fourth Power of Paths 111
[3] J.Devaraj and P.Delphy, On signed cordial graph, Int. J. of Mathematical Sciences and
Applications, Vol.1, No.3, September (2011), 1156-1167.
[4] A .T. Diab, On cordial labeling of the second power of paths with other graphs, Ars
Combinatoria, 97A(2010), 327-343.
[5] A. Elrokh and A. Abd El-hay, The cordiality of cone and lemniscate graphs, MJMS (sub-
mitted).
[6] J. A. Gallian, A dynamic survey of graph labeling, The Electronic Journal of Combina-
torics, 17(2010), DS6.
[7] S. W. Golomb, How to number a graph, in Graph Theory and Computing, R.C. Read ed.,
Academic Press, New York, 1972, 23-37.
[8] R. L. Graham, and N. J. A. Sloane, On additive bases and harmonious graphs, SIAM J.
Alg. Discrete Math., 1(1980), 382-404.
[9] F.Harray, On the notion of a signed graph, Michigan Math. J., 2 (1953-1954), 143-146.
[10] M.Hovey, A-cordial Graphs, Discrete Mathematics, 93 (1991), 183-194.
[11] S. Nada, A. Elrokh and A. Abd El-hay. The cordiality of the second power of some graphs,
CHZ.Math Journal(submitted).
[12] A. Rosa, On certain valuations of the vertices of a graph, Theory of Graphs(Internat
Symposium, Rome, July 1966), Gordon and Breach, N.Y. and Dunod Paris (1967), 349-
355.
[13] M. A. Seoud, A. T. Diab, and E. A. Elsahawi, On strongly-C harmonious, relatively prime,
odd graceful and cordial graphs, Proc. Math. Phys. Soc. Egypt, 73(1998), 33-55.
[14] T.Zaslausky, Signed graphs, Discrete applied Mathematics, 4(1982), 47-74.
International J.Math. Combin. Vol.4(2019), 112-121
On the Eccentric Sequence of Composite Graphs
Yaser Alizadeh
(Department of Mathematics, Hakim Sabzevari University, Sabzevar, Iran)
E-mail: [email protected]
Abstract: The eccentric sequence of a graph is defined as list of eccentricity of its vertices.
Eccentric sequence of composite graphs under seven graph products: line graph, sum, carte-
sian product, disjunction, symmetric difference, lexicographic product and corona product
is investigated. Also some family of non vertex transitive graphs that are self centered are
determined as product of graphs. It is proved that for any positive integer d, there is an
infinite family of non-vertex transitive self centered graphs with diameter d. The relation
between total eccentricity of a tree and total eccentricity of its line graph is given.
Key Words: Eccentricity, eccentric sequence, graph product, self centered graph.
AMS(2010): 05C76.
§1. Introduction
We consider only simple connected graphs in this paper. Let G = (V (G), E(G)) be a graph and
u, v be two vertices of G. The distance between u and v, dG(u, v) (simply d(u, v)) is the length
of shortest path connecting u and v. For a vertex v ∈ V (G), the eccentricity of v, εG(v) is the
maximum distance from v to other vertices in G. The maximum and the minimum eccentricity
among all vertices of G are called diameter diam(G) and radius rad(G) of G respectively. The
center of G, C(G) is the set of vertices whose eccentricity is equal to rad(G). A graph G is
called self centered if all of its vertices have a same eccentricity. Let V (G) = {v1, v2, · · · , vn}be the vertex set of G. Eccentric sequence of G is the sequence ε1, ε2, · · · , εn where εi is
the eccentricity of vertex vi. The minimum eccentric sequence of G, es(G) is the sequence
{εt11 , ε
t22 , · · · , εtk
k } where ε1, ε2 · · · , εk are the different eccentricities of vertices and ti denotes
the number of vertices with eccentricity εi and more over εi+1 = εi + 1 for 1 6 i 6 k − 1.
Note that ε1 = rad(G) and εk = diam(G). Eccentric sequence is interesting since it provides
information on the vertex eccentricities and some structural properties of the graph such as
diameter, radius and variability of vertex eccentricities. Call a sequence of positive integer
eccentric if it is eccentric sequence of a graph. In a series of papers several properties of
eccentric sequences are studied. For instance see surveys [3, 7, 11, 14, 17]. Characterization of
eccentric sequences of graphs was first considered by Lesniak [11] who characterized sequences
which are the eccentricity sequences of trees.
1Received July 23, 2019, Accepted December 8, 2019.
On the Eccentric Sequence of Composite Graphs 113
Study of graph invariants specially topological indices under graph products is very in-
terested in mathematical literature. Some properties and application are reported in surveys
[1, 2, 4 C 6, 8 C 10, 12, 15, 18]. In this paper, we study the eccentric sequence of composite
graphs. We obtain explicit formulas of eccentric sequence for some graph product such as: line
graph, sum, cartesian product, disjunction, symmetric difference, lexicographic product and
corona product. Two important topological indices based on eccentricity of vertices are the
total eccentricity and eccentric connectivity index. The total eccentricity of a graph G, ξ(G) is
the sum of eccentricities of its vertices. Clearly, if
es(G) = {εt11 , · · · , εtk
k }
then,
ξ(G) =k∑
i=1
tiε(vi).
The eccentric connectivity index of graph G, ECI(G), introduced by Sharma et al. [16], is
defined as
ECI(G) =∑
v∈V (G)
ε(v)deg(v),
where deg(v) denotes degree of vertex v. These topological indices have been used as mathe-
matical models for the prediction of biological activities of diverse nature. The automorphism
group of G is denoted with Aut(G). A graph is called vertex transitive if for any pair of vertices
u and v, there is an automorphism α such that α(u) = v. It is known that an automorphism
of a graph preserve the distance function. It follows that vertex-transitive graphs are always
regular and self centered graph.
In this paper we construct an infinite family of non-vertex transitive graphs which are
self centered. In the rest of the section, some standard graph products are introduced, then
eccentric sequence of graphs under these graph products is verified. First, we start with line
graph. Line graph of G, L(G) is a graph which each vertex of L(G) is associated with an edge
of G and two vertices in L(G) is adjacent if and only if the corresponding edges of G have
an end vertex in common. The sum of two graphs G1 and G2, G1 + G2 is defined as the
graph with the vertex set V (G1) ∪ V (G2) and the edge set E (G1 +G2) = E (G1) ∪ E (G2) ∪{u1u2 |u1 ∈ V (G1), u2 ∈ V (G2)} . The next binary graph product is cartesian product. The
Cartesian product G1�G2 is the graph with vertex set V (G1)× V (G2) and (u1, u2) is adjacent
to (v1, v2) if u1 = v1 and (u2v2) ∈ E(G2), or u2 = v2 and (u1v1) ∈ E(G1).
The disjunction G1 ∨G2 is the graph with vertex set V (G1) × V (G2) and
E (G1 ∨G2) = {(u1, u2)(v1, v2) |u1v1 ∈ E(G1) or u2v2 ∈ E(G2) } .
For given graphs G1 and G2, their symmetric difference G1
⊕G2 is the graph with vertex
set V (G1)×V (G2) and two vertices (u1, u2) and (v1, v2) are adjacent if and only if u1v1 ∈ E(G1)
or u2v2 ∈ E(G2)
The diameter of disjunction and symmetric difference of two graphs when both of them
114 Yaser Alizadeh
contain more than one vertex do not exceed of 2. The next binary operation is the lexicographic
product. The lexicographic product of two graphs G1 and G2, G1 [G2] is the graph with vertex
set V (G1) × V (G2) and two vertices u = (u1, u2) and v = (v1, v2) are adjacent if (u1 is
adjacent with v1) or (u1 = v1 and u2 and v2 are adjacent).The operations sum, disjunction and
symmetric difference are symmetric operation and this fact implies that they have symmetric
eccentric sequence. But the lexicographic product do not have such property. Let ni, i = 1, 2
denotes the order of Gi. The corona product of two graphs is denoted by G1◦G2 and is obtained
from one copy of G1 and n1 copies of G2, and then joining all vertices of the i-th copy of G2 to
the i-th vertex of G1 for i = 1, 2, · · · , n1. In [13], application of coronas in chemical modeling
was reported.
§2. Main Result
In this section, explicit formulas for eccentric sequence of some composite graphs is given.
2.1 Line Graph of Trees
Eccentric sequence of line graph of a tree can be determined by its eccentric sequence. We
present a relation between of total eccentricity of a tree and total eccentricity of its line graph.
Theorem 2.1 Let T be a tree with rad(T) = r. If es(T ) = {rn0 , (r + 1)n1 , · · · , (2r)nr}, then
eccentric sequence of L(T ) is obtained as
es(L(T )) =
{rn1 , (r + 1)n2 , · · · , (2r − 1)nr} if C(T ) = K1
{(r − 1)1, rn1 , · · · , (2r − 2)nr−1} if C(T ) = K2
where nr > 0 and for 0 6 i 6 r − 1, ni > 1.
Proof We must to consider two cases.
Case 1. C(T ) = K1.
Let p be the unique central vertex. Let Ni(p) = {v ∈ V (T )|d(p, v) = i}. Since T is a tree,
for a vertex u ∈ Ni(p), εT (u) = i+ r. Also there is a unique vertex v ∈ Ni−1(p) that is adjacent
to u. Now consider the bijection f : V (T ) − {p} → E(T ), if u ∈ Ni(p), i > 1, then f(u) = uv
where v ∈ Ni−1(p). For a vertex u ∈ Ni(p), we have εT (u) = r+ i and εL(T )(f(u)) = r+ i− 1.
Thus if
es(T ) = {r1, (r + 1)n1 , · · · , (2r)nr},
the eccentric sequence of L(T ) is obtained as
es(L(T )) = {rn1 , (r + 1)n2 , . . . , (2r − 1)nr}.
Case 2. C(T ) = K2.
On the Eccentric Sequence of Composite Graphs 115
Let p1 and p2 be two adjacent central vertices of T . It is easy to see εL(T )(p1p2) = r − 1.
Let T−{p1p2} = T1∪T2 which pj ∈ Tj for j = 1, 2. Clearly if u ∈ Tj, clearly ε(u) = r+d(u, pj).
By a similar argument to Case 1, if u ∈ Ni(Pj)∩ Tj , j = 1, 2, then there is a unique vertex v ∈Ni−1(Pj) that uv ∈ E(T ). Thus we get again a bijection f : V (T )−{p1, p2} → E(T )−{p1p2}.Also if u ∈ Ni(Pj) ∩ Tj, then εL(T )(f(u)) = εT (u) − 1. This implies that if
es(T ) = {r2, (r + 1)n1 , · · · , (2r − 1)nr−1},
then
es(L(T )) = {(r − 1)1, (r)n1 , · · · , (2r − 2)nr−1}. 2Corollary 2.2 Let T be a tree. L(T ) is self centered graph if and only if T is a star graph.
Proof Clearly the line graph of a star graph is complete graph and then is self centered.
Let T be a tree of order n and rad(T) = r. If L(T ) is self center graph, by Theorem 2.1,
eccentric sequence of T has form es(T ) = {(r − 1)n−1} or {r1, (r + 1)n−2}. This means that
the variability of eccentricity in T is 1 or 2. Since T is a tree, this implies that rad(T) = 1 and
the proof is completed. 2Corollary 2.3 Let T be a tree of order n and radius r. Then
ξ(T ) = ξ(L(T )) + n+ r − 1.
Proof We consider two cases with respect to es(T ) and es(L(T )).
Case 1. es(T ) = {r1, (r + 1)n1 , · · · , (2r)nr} and es(L(T )) = {rn1 , (r + 1)n2 , · · · , (2r − 1)nr}.
By a straight calculation we get
ξ(T ) − ξ(L(T )) = r +
r∑
i=1
ni = r + n− 1.
Case 2. es(T ) = {r2, (r+1)n1 , · · · , (2r−1)nr−1} and es(L(T )) = {r − 11, (r)n1 , · · · , (2r−2)nr−1}.
Again, in this case a same result is obtained as well.
ξ(T ) − ξ(L(T )) = 2r − (r − 1) +
r−1∑
i=1
ni = r + 1 + n− 2 = n+ r − 1. 22.2 Sum
Theorem 2.4 Let G1 and G2 be simple connected graphs. Then,
es(G1 +G2) = {1c1+c2 , 2n1+n2−c1−c2},
116 Yaser Alizadeh
where ci is the number of vertices of eccentricity 1 in Gi and ni is the number of vertices of
Gi; i = 1, 2.
Proof It is not difficult to see that diam(G1 + G2) 6 2. For vertex x ∈ V (Gi) we have
εG1+G2(x) = 1 if and only if εGi(x) = 1. Let ci denotes the number vertices of eccentricity
1 in Gi, i = 1, 2. Then the eccentric sequence of G1 + G2 is obtained as es(G1 + G2) =
{1c1+c2 , 2n1+n2−c1−c2}. 2The eccentric sequence of sum of more than two graphs can be obtained by a reasoning
similar to the above.
Corollary 2.5 Let G1, G2, · · · , Gk be simple connected graphs. Then
es(G1 +G2 + · · · +Gk) = {1∑
ki=1 ci , 2
∑ki=1 ni−ci},
where ci is the number of vertices of eccentricity 1 (or 0) in Gi and ni is the number of vertices
of Gi; i = 1, 2, · · ·n.
Corollary 2.6 For any integer n > 5, there is a self centered graph and non vertex transitive
of diameter 2 and order n.
Proof It is sufficient to consider the complete bipartite graph K2,n−2 = K2 + Kn−2. 22.3 Cartesian Product
Theorem 2.7 Let es(G1) = {ε1t1 , ε2t2 , · · · , εk
tk} and es(G2) = {δ1s1 , δ2s2 , · · · , δmsm}.
Then,
es(G1�G2) ={
(εi + δj)tisj
}
16i6k, 16j6m.
Proof It is known that dG1�G2((x, y), (u, v)) = dG1(x, u) + dG2(y, v), this implies that
εG1�G2(x, y) = εG1(x) + εG2(y).
Let mi and nj be the number of vertices of eccentricity εi and δj in G1 and G2 re-
spectively. Then, minj vertices of G1�G2 have eccentricity εi + δj . Therefore es(G1�G2) ={(εi + δj)
tisj
}
16i6k, 16j6m. 2
Corollary 2.8 There are infinite family of non-vertex transitive self centered graph.
Proof It is sufficient to consider the powers of a non-vertex transitive self centered graph
such as Km,n where m 6= n and m,n > 2. 22.4 Disjunction
First note that if G = K1 then G ∨H ∼= H and G⊕H ∼= H as well. Therefore the considered
graph for these two graph products are except K1.
On the Eccentric Sequence of Composite Graphs 117
Theorem 2.9 Let G1 6= K1 6= G2. Then
es(G1 ∨G2) = {1c1c2 , 2n1n2−c1c2}
where ci is the number of vertices of eccentricity 1 in Gi and ni is the number of vertices of
Gi; i = 1, 2.
Proof Let (x, y) and (u, v) be two vertices of G1 ∨G2 and xx′ ∈ E(G1) and vv′ ∈ E(G2).
Since (x, y) and (u, v) both are adjacent to (x′, v′) then d((x, y), (u, v)) 6 2. Therefore for each
vertex (x, y) ∈ G1 ∨ G2, we have εG1∨G2(x, y) 6 2. If εG1(x) > 1, and dG1(x, y) > 2 then
d((x, u), (y, u)) = 2. Hence, εG1∨G2(x, y) = 1 if and only if εG1(x) = 1 = εG2(y). Let ci vertices
of Gi are of eccentricity 1 for i = 1, 2. Then c1c2 vertices of G1 ∨G2 have eccentricity 1 and the
other vertices are of eccentricity 2. This implies that es(G1 ∨G2) = {1c1c2 , 2n1n2−c1c2}. 2Corollary 2.10 Let G1 and G2 be two graphs of radius at least 2. Then G1 ∨ G2 is a self
centered graph and es(G1 ∨G2) = {2n1n2}.
2.5 Symmetric Difference
Lemma 2.11([9]) Let G1 and G2 be two simple connected graphs. The number of vertices
of Gi is denoted by ni for i = 1, 2. Then degG1⊕G2((u, v)) = n2degG1(u) + n1degG2(v) −2degG1(u)degG2(v).
Theorem 2.12 Let G1 6= K1 6= G2. Then es(G1 ⊕G2) = {2n1n2}.
Proof Let (x, y) and (u, v) be two vertices of G1 ∨G2 and xx′ ∈ E(G1) and vv′ ∈ E(G2).
Since (x, y) and (u, v) are adjacent to (x, v′) then d((x, y), (u, v)) 6 2. On the other hand,
(x, y) and (x′, v′) are not adjacent in G1 ⊕G2. Therefore each vertex (x, y) ∈ G1 ⊕G2 we have
εG1⊕G2(x, y) = 2. This concludes that es(G1 ⊕G2) = {2n1n2}. 2Corollary 2.13 For any positive integer d there is an infinite family of non vertex transitive
graphs which are self centered with diameter d.
Proof Let ⊕ki=1Gi = G1 ⊕ G2 ⊕ · · · ⊕ Gk. For any positive integers n, k > 3, let Gn,k =
⊕ki=1Pn. Using Lemma 2.11 and Theorem 2.12 we get that Gn,k is a self centered graph of
diameter 2 and it is non vertex transitive because it is not regular. Now consider the graph
Hn,k,d = Gn,k�C2(d−2) which is self center graph of diameter d. Since Gn,k is not regular
graph then Hn,k,d is not regular and consequently is not vertex transitive graph as well. Clearly
diameter of Hn,k,d is d and the proof is completed. 22.6 Lexicographic Product
For the lexicographic product of graphs, the distance of pair vertices is determined by the
following lemma.
118 Yaser Alizadeh
Lemma 2.14([9]) The distance of pair vertices in G1[G2] is
dG1[G2] ((u1, v1), (u2, v2)) =
dG1(u1, u2) v1 = v2
1 u1 = u2, v1v2 ∈ E(G2)
2 otherwise
Therefore, the eccentricity of vertex (u, v) ∈ V (G1[G2]) is determined by
ε(u, v) =
1 if εG1(u) = εG2(v) = 1
2 if εG1(u) = 1 and εG2(v) > 2
εG1(u) if ε(u) > 2
Now, all conditions are ready to obtain the eccentric sequence of G1[G2].
Theorem 2.15 Let es(G1) = {εt11 , . . . , ε
tk
k } and ni and ci, i = 1, 2 be the order and the number
of vertices of eccentricity 1 of Gi respectively. Then
es(G1[G2]) = {1c1c2 , 2c1(n2−c2)+t2n2 , εt3n23 , · · · , εtkn2
k }.
2.7 Corona
Remark 2.16 If G1 = K1, then G1oG2 = K1 +G2 and
es(G1oG2) = {11+c2 , 2n2−c2}
.
Theorem 2.17 Let G1 6= K1 and es(G1) ={εti
i
}k
i=1. Then
es(G1oG2) ={εt12 , ε
t2+n2t13 , εt3+n2t2
4 , · · · , εtk+n2tk−1
k+1 , εn2tk
k+2
},
where n2 = |V (G2)| and εi+1 = εi + 1.
Proof Let V (G1) = {v1, v2, . . . , vn} and G2,i be the copy of G2 associated to vi. From
the structure of corona product of graphs one can see that εG1o G2(vi) = εG1(vi) + 1 and if
x ∈ V (G2,i), εG1oG2(x) = εG1(vi) + 2, 1 6 i 6 n. Then for x ∈ V (G1oG2),
ε2 = ε1 + 1 6 ε(x) 6 εk + 2 = εk+2.
Let v be a central vertex of G1 and εG1(v) = ε1, then εG1oG2(v) = ε1 +1 = ε2. This follows that
center of G1 coincides center of G1oG2. For i > 2, the set of vertices of G1 having eccentricity
εi and the vertices of G2,t which ε(Vt) = εi−1 are of eccentricity εi+1 in G1oG2. Thus for i > 2,
the number of vertices that have eccentricity εi+1 is ti + n2ti−1. The proof is completed. 2
On the Eccentric Sequence of Composite Graphs 119
§3. Examples and Concluding Remarks
In this section, our theorems for eccentric sequence are illustrated for several more particular
composite graphs. We first give the expressions for suspensions.
Corollary 2.18 Let G be a graph on n vertices. Then
es(K1 +G) = {1c+1, 2n−c},
where c is the number of vertices of eccentricity 1 in G.
Next, the eccentric sequence for the fan graph K1 +Pn and the wheel graph Wn = K1 +Cn
are presented by
Corollary 2.19 For the fan graph K1 + Pn and the wheel graph Wn = K1 + Cn,
es(K1 + Pn) =
{12} if n = 1,
{13} if n = 2,
{12, 22} if n = 3,
{11, 2n} if n > 4
and
es(Wn) =
{14} if n = 3
{11, 2n} if n > 4.
By composing paths and cycles with various small graphs, we can obtain different classes
of polymer like graphs. For example, we state the eccentric sequence for the fence graph Pn[K2]
and the closed fence Cn[K2] in the following conclusion.
Corollary 2.20 For the fence graph Pn[K2] and the closed fence Cn[K2],
es(Pn[K2]) =
{12} if n = 1,
{14} if n = 2,
{12, 24} if n = 3,
{22t2 , 32t3 , · · ·k2tk} if n > 4,
where es(Pn) = {2t2, 3t3 , · · · ktk} for n > 4 and
es(Cn[K2]) =
{16} if n = 3,
{[n2 ]2n} if n > 4.
The t-thorny graph of a given graph G is obtained as GoKn, where Kn denotes the empty
graph on n vertices. For the t-thorny path and t-thorny cycle we get the following eccentric
120 Yaser Alizadeh
sequence.
Corollary 2.21 For the t-thorny graph PnoKt,
es(PnoKt) =
{11, 2t} n = 1,
{22, 32t} n = 2,
{21, 3t+2, 42t} n = 3.
If n > 4 and it is even
es(PnoKt) = {(n2
+ 1)2, (n
2+ 2)2t+2, · · · , n2t+2, (n+ 1)2t}
and if n > 5 and it is odd then
es(PnoKt) = {(n+ 1
2)1, (
n+ 1
2+ 1)t+2, (
n+ 1
2+ 2)2t+2, · · · , n2t+2, (n+ 1)2t}.
References
[1] Alizadeh Y., Iranmanesh A., Doslic T., Additively weighted Harary index of some com-
posite graphs, Discrete Math., 313 (2013), 26-34.
[2] Azari M., Iranmanesh A. Joins, coronas and their vertex-edge Wiener polynomials, Tamkang
J.Math., 47 (2016), 163–178.
[3] Behzad M., Simpson J. E., Eccentric sequences and eccentric sets in graphs, Discrete
Math., 16 (1976), 187–193.
[4] Doslic T., Saheli M., Eccentric connectivity index of composite graphs, Util. Math. 95
(2014), 3–22.
[5] Eskender B., Vumar E., Eccentric connectivity index and eccentric distance sum of some
graph operations, Trans. Combin. 2 (2013), 103–111.
[6] Hammack R., Imrich W., Klavzar S., Handbook of Product Graphs, CRC Press, Boca
Raton, 2011.
[7] Haviar A., Hrnciar P., Monoszov G., Minimal eccentric sequences with Least eccentricity
three, Acta Univ. M. Belii. Math. 5 (1997), 27-50.
[8] Imrich W., Klavzar S., Product Graphs: Structure and Recognition, John Wiley & Sons,
New York, USA, 2000.
[9] Khalifeh M.H., Yousefi-Azari H., Ashrafi A.R. The hyper-Wiener index of graph opera-
tions, Comput. Math. Appl. 56 (2008), 1402-1407.
[10] Khalifeh H., Youseri-Azari H., Ashrafi A. R., Vertex and edge PI indices of Cartesian
product of graphs, Discrete Appl. Math. 156 (2008), 1780–1789.
[11] Lesniak, L. Eccentric sequences in graphs, Period. Math. Hung. 6 (1975), 287–293.
[12] Mamut A., Vumar E., Vertex vulnerability parameters of Kronecker products of complete
graphs, Inform. Process. Lett. 106 (2008), 258–262.
[13] Milicevic A., Trinajstic N., Combinatorial enumeration in chemistry, in Chemical Mod-
On the Eccentric Sequence of Composite Graphs 121
elling: Applications and Theory, A. Hincliffe, Ed. Cambridge: RSC Publishing, 2006,
405–469.
[14] Nandakumar, R. On Some Eccentric Properties of Graphs, Ph.D. Thesis, Indian Institute
of Technology, India (1986).
[15] Pattabiraman K., Vijayaragavan M., Reciprocal degree distance of some graph operations,
Trans. Comb. 2 (2013), 13–24.
[16] Sharma V., Goswami R., Madan A.K. Eccentric connectivity index: A novel highly dis-
criminating topological descriptor for structure-property and structure-activity studies, J.
Chem. Inf. Comput. Sci. 37 (1997), 273–282.
[17] Skurnik R., Eccentricity sequences of trees and their caterpillars cousins, Graph Theory
Notes of New York, 37 (1999), 20-24.
[18] Yarahmadi Z., Moradi S., The center and periphery of composite graphs, Iranian J. Math.
Chem., 5 (2014), 35–44.
International J.Math. Combin. Vol.4(2019), 122-135
F -Centroidal Mean Labeling of Graphs Obtained From Paths
S. Arockiaraj1, A. Rajesh Kannan2 and A. Durai Baskar3
1. Department of Mathematics, Government Arts & Science College, Sivakasi- 626124, Tamil Nadu, India
2. Department of Mathematics, Mepco Schlenk Engineering College, Sivakasi- 626005, Tamil Nadu, India
3. Department of Mathematics, C.S.I. Jayaraj Annapackiam College, Nallur - 627853, Tamil Nadu, India
E-mail: [email protected], [email protected], [email protected]
Abstract: A function f is called an F -centroidal mean labeling of a graph G(V, E) with p
vertices and q edges if f : V (G) → {1, 2, 3, · · · , q + 1} is injective and the induced function
f∗ : E(G) → {1, 2, 3, · · · , q} defined as
f∗(uv) =
⌊2 [f(u)2 + f(u)f(v) + f(v)2]
3 [f(u) + f(v)]
⌋,
for all uv ∈ E(G), is bijective. A graph that admits an F -centroidal mean labeling is called
an F -centroidal mean graph. In this paper, we have discussed the F -centroidal meanness of
the graph Pn(X1, X2, · · ·Xn), the twig graph TW (Pn), the graph Pn ◦Sm for m ≤ 4, planar
grid Pm × Pn for m ≤ 3, the ladder graph Ln, the graph Pn ◦ K2, the graph P ba for a ≥ 2
and b ≤ 3, the middle graph of the path, total graph of the path and the square graph of
the path, the splitting graph of the path and the graph P (1, 2, · · · , n − 1).
Key Words: Labeling, F -centroidal mean labeling, F -centroidal mean graph, Smaran-
dachely F -centroidal mean labeling.
AMS(2010): 05C78.
§1. Introduction
Throughout this paper, by a graph we mean a finite, undirected and simple graph. Let G(V,E)
be a graph with p vertices and q edges. For notations and terminology, we follow [8]. For a
detailed survey on graph labeling, we refer [7].
Path on n vertices is denoted by Pn. The graph Pn(X1, X2, . . . Xn), is a tree obtained
from a path on n vertices by attaching Xi pendent vertices at each ith vertex of the path, for
1 ≤ i ≤ n. A Twig TW (Pn), n ≥ 4 is a graph obtained from a path by attaching exactly two
pendant vertices to each internal vertices of the path Pn. The graph G ◦Sm is obtained from G
by attaching m pendant vertices to each vertex of G. Let G1 and G2 be any two graphs with
p1 and p2 vertices respectively. Then the Cartesian product G1 × G2 has p1p2 vertices which
are {(u, v) : u ∈ G1, v ∈ G2} and the edges are obtained as follows: (u1, v1) and (u2, v2) are
adjacent in G1×G2 if either u1 = u2 and v1 and v2 are adjacent in G2 or u1 and u2 are adjacent
1Received January 11, 2019, Accepted December 9, 2019.
F -Centroidal Mean Labeling of Graphs Obtained From Paths 123
in G1 and v1 = v2. The product Pm × Pn and is called a planar grid and P2 × Pn is called a
ladder, denoted by Ln.
Let a and b be integers such that a ≥ 2 and b ≥ 2. Let y1, y2, . . . ya be the ′a′ fixed
vertices. Connect yi and yi+1 by means of b internally disjoint paths P ji of length ′i+ 1′ each,
for 1 ≤ i ≤ a− 1 and 1 ≤ j ≤ b. The resulting graph embedded in the plane is denoted by P ba .
The middle graph M(G) of a graph G is the graph whose vertex set is {v : v ∈ V (G)} ∪ {e :
e ∈ E(G)} and the edge set is {e1e2 : e1, e2 ∈ E(G) and e1 and e2 are adjacent edges of
G} ∪ {ve : v ∈ V (G), e ∈ E(G) and e is incident with v}. The total graph T (G) of a graph G is
the graph whose vertex set is V (G) ∪ E(G) and two vertices are adjacent if and only if either
they are adjacent vertices of G or adjacent edges of G or one is a vertex of G and the other
one is an edge incident on it. Square of a graph G, denoted by G2, has the vertex set as in G
and two vertices are adjacent in G2 if they are at a distance either 1 or 2 apart in G. For each
vertex v of the graph G, take a new vertex v′ to these vertices of G adjacent to v. The graph
thus obtained is called the splitting graph G and it is denoted by S′(G). An arbitrary super
subdivision P (m1,m2, · · · ,mn−1) of a path Pn is a graph obtained by replacing each ith edge
of Pn by identifying its end vertices of the edge with a partition of K2,mihaving 2 elements,
where mi is any positive integer.
Durai Baskar and Arockiaraj defined the F -harmonic mean labeling [6] and discussed its
meanness of some standard graphs. The concept of F -geometric mean labeling was introduced
by Durai Baskar and Arockiaraj [5] and it was developed [4]. The concept of F -root square
mean labeling was introduced by Arockiaraj et al., [1] and they studied the F -root square
mean labeling of some standard graphs [2]. Durai Baskar and Manivannan were introduced
F -heronian mean labeling [3]. Motivated by the works of so many authors in the area of graph
labeling, we introduced a new type of labeling called an F -centroidal mean labeling.
A function f is called an F -centroidal mean labeling of a graph G(V,E) with p vertices
and q edges if f : V (G) → {1, 2, 3, · · · , q+1} is injective and the induced function f∗ : E(G) →{1, 2, 3, · · · , q} defined by
f∗(uv) =
⌊2 [f(u)2 + f(u)f(v) + f(v)2]
3 [f(u) + f(v)]
⌋,
for all uv ∈ E(G), is bijective. Otherwise, it is called a Smarandachely F -centroidal mean
labeling of G if there is a number k ∈ {1, 2, 3, · · · , q} such that the inverse f−∗ of f∗ holds with
|f−∗(k)| ≥ 2. A graph that admits an F -centroidal mean labeling is called an F -centroidal
mean graph.
An F -centroidal mean labeling of cycle C4 is given in Figure 1.rrr r2
1
1 2 4
3 5
4
Figure 1 An F -centroidal mean labeling labeling of C4
124 S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar
In this paper, we have discussed the F -centroidal meanness of the graph Pn(X1, X2, · · ·Xn),
the twig graph TW (Pn), the graph Pn ◦ Sm for m ≤ 4, planar grid Pm × Pn for m ≤ 3, the
ladder graph Ln, the graph Pn ◦K2, the graph P ba for a ≥ 2 and b ≤ 3, the middle graph of the
path, total graph of the path and the square graph of the path, the splitting graph of the path
and the graph P (1, 2, · · · , n− 1).
§2. Main Results
Theorem 2.1 The graph Pn(X1, X2, · · ·Xn) is an F -centroidal mean graph, for 1 ≤ Xi ≤ 3
and |Xi −Xi+1| ≤ 1, for 1 ≤ i ≤ n− 1.
Proof Let u1, u2, · · · , un be the vertices of the path Pn. Let v(1)i , v
(2)i , · · · , v(Xi)
i be the
pendant vertices attached at ui, for 1 ≤ i ≤ n.
Define f : V (Pn(X1, X2, · · ·Xn)) → {1, 2, 3, · · · ,n∑
i=1
Xi + n} as follows:
f(v(1)i ) =
2, X1 = 1,
1, X1 6= 1.
For 2 ≤ i ≤ n,
f(v(1)i ) =
i−1∑k=1
Xk + i, Xi = 2, 3,
i−1∑k=1
Xk + i+ 1, Xi = 1.
For 1 ≤ i ≤ n,
f(v(j)i ) =
f(v
(1)i ) + 2, j = 2
f(v(1)i ) + 3, Xi = 3 and j = 3
and
f(ui) =
f(v(1)i ) + 1, Xi = 2, 3,
f(v(1)i ) − 1, Xi = 1.
Then the induced edge labeling f∗ is obtained as follows:
For 1 ≤ i ≤ n− 1,
f∗(uiui+1) =
f(ui) + 1, Xi = 1, 2,
f(ui) + 2, Xi = 3
and f∗(v(1)1 u1) = 1.
F -Centroidal Mean Labeling of Graphs Obtained From Paths 125
For 1 ≤ i ≤ n,
f∗(v(1)i ui) =
f(v
(1)i ) + 1, Xi = 2, 3,
f(v(1)i ) − 1, Xi = 1
and
f∗(v(j)i ui) =
f(ui), Xi = 2, 3 and j = 2,
f(ui) + 1, Xi = 3 and j = 3.
Hence f is an F -centroidal mean labeling of Pn(X1, X2, · · ·Xn).Thus the graph Pn(X1, X2, · · ·Xn)
is an F -centroidal mean graph, for 1 ≤ Xi ≤ 3 and |Xi −Xi+1| ≤ 1, for 1 ≤ i ≤ n− 1. 21 3 4 6 8 9 11 12 14 15 16 18 19 20 22
2 5 7 10 13 17 213 6 8 11 15 19
1 2 4 5 7 9 10 1213
14 16 18 20 2117
s s s s s sss s s s s s s s s s s s s s sFigure 2 An F -centroidal mean labeling of Pn(2, 2, 1, 2, 3, 3, 2)
Corollary 2.2 The twig graph TW (Pn) of the path Pn is an F -centroidal mean graph, for
n ≥ 4.
Theorem 2.3 The graph Pn ◦ Sm is an F -centroidal mean graph, for n ≥ 1 and m ≤ 4.
Proof Let v1, v2, v3, · · · , vn be the vertices of the path Pn and u(i)1 , u
(i)2 , u
(i)3 , · · · , u(i)
m be
the pendant vertices at each vi, for 1 ≤ i ≤ n.
Case 1. m = 4.
Define f : V (Pn ◦ S4) → {1, 2, 3, . . . , 5n} as follows:
f(v1) = 2,
f(vi) = 5i− 2, for 1 ≤ i ≤ n,
f(u(1)1 ) = 1,
f(u(i)1 ) = 5i− 5, for 2 ≤ i ≤ n,
f(u(1)2 ) = 3,
f(u(i)2 ) = 5i− 3, for 2 ≤ i ≤ n,
f(u(i)3 ) = 5i− 1, for 1 ≤ i ≤ n,
f(u(i)4 ) = 5i+ 1, for 1 ≤ i ≤ n− 1,
f(u(n)4 ) = 5n.
126 S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar
Then the induced edge labeling f∗ is obtained as follows:
f∗(vivi+1) = 5i, for 1 ≤ i ≤ n− 1,
f∗(viu(i)1 ) = 5i− 4, for 1 ≤ i ≤ n,
f∗(viu(i)2 ) = 5i− 3, for 1 ≤ i ≤ n,
f∗(viu(i)3 ) = 5i− 2, for 1 ≤ i ≤ n
f∗(viu(i)4 ) = 5i− 1, for 1 ≤ i ≤ n.
Case 2. 1 ≤ m ≤ 3.
By Theorem 2.1, the results follows in this case.
Hence f is an F -centroidal mean labeling of Pn ◦Sm, for n ≥ 1 and m ≤ 4. Thus the graph
Pn ◦ Sm is an F -centroidal mean graph, for n ≥ 1 and m ≤ 4. 2sTTTTTs s ss TTTTTs s ss TTTTTs s ss TTTTTs s sss2 8 13 18
3 4 6 5 7 9 11 10 12 14 16 15 17 19 20
5 10 15
12 3
4 67 8
9 1112 13
1417
16 19
18
s s1
Figure 3 An F -centroidal mean labeling of P4 ◦ S4
Theorem 2.4 The planar grid Pm ×Pn, is an F -centroidal mean graph, for m ≤ 3 and n ≥ 2.
Proof Let V (Pm × Pn) = {vij : 1 ≤ i ≤ m, 1 ≤ j ≤ n} and E(Pm × Pn) = {vijv(i+1)j :
1 ≤ i ≤ m− 1, 1 ≤ j ≤ n}∪ {vijvi(j+1) : 1 ≤ i ≤ m, 1 ≤ j ≤ n− 1} be the vertex set and edge
set of the graph Pm × Pn.
Case 1. m = 2.
Define f : V (P2 × Pn) → {1, 2, 3, · · · , 3n− 1} as follows:
f(vij) = i+ 3(j − 1), for 1 ≤ i ≤ 2 and 1 ≤ j ≤ n.
Then the induced edge labeling f∗ is obtained as follows:
f∗(v1jv2j) = 3j − 2, for 1 ≤ i ≤ n,
f∗(vijvi(j+1)) = i+ 3j − 2, for 1 ≤ i ≤ 2 and 1 ≤ j ≤ n− 1.
Case 2. m = 3.
F -Centroidal Mean Labeling of Graphs Obtained From Paths 127
Define f : V (P3 × Pn) → {1, 2, 3, · · · , 5n− 2} as follows:
f(vi1) = i, for 1 ≤ i ≤ 3,
f(vi2) =
i+ 4, i = 1,
i+ 5, 2 ≤ i ≤ 3,
f(vij) = i+ 5(j − 1), for 1 ≤ i ≤ 3 and 3 ≤ j ≤ n.
Then the induced edge labeling f∗ is obtained as follows:
f∗(vi1v(i+1)1) = i, for 1 ≤ i ≤ 2,
f∗(vi1vi2) = i+ 2, for 1 ≤ i ≤ 3,
f∗(vijv(i+1)j) =
2i+ 3(j − 1), 1 ≤ i ≤ 2 and j = 2,
i+ 5(j − 1), 1 ≤ i ≤ 2 and 3 ≤ j ≤ n,
f∗(vijvi(j+1)) = i+ 5j − 3, for 1 ≤ i ≤ 3 and 2 ≤ j ≤ n− 1.
Hence the graph Pm × Pn admits an F -centroidal mean labeling. Thus the graph Pm × Pn is
an F -centroidal meangraph for m ≤ 3.
For n = 2, an F -centroidal mean labeling of P2 × P4 is as shown in Figure 4. 2tttttttt12 4 5 7 8 10
10
11986532
1 4 7
Figure 4 An F -centroidal mean labeling of P2 × P4t t t ttt t t t ttt t t t ttt
51 11 16 21 26 31tt
3 8 13 18 23 28 33
27 12 17 22 27
32t13
6
4
2 75
8
11
12
9
10
1316
14
1517
1821
19
2022
2326
24
25
27
2831
29
30
32
Figure 5 An F -centroidal mean labeling of P3 × P7
Corollary 2.5 Every Ladder graph Ln = P2 × Pn is an F -centroidal mean graph for n ≥ 2.
Theorem 2.6 The graph Pn ◦K2 is an F -centroidal mean graph for n ≥ 1.
128 S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar
Proof Let v1, v2, v3, · · · , vn be the vertices of the path Pn and u(1)i , u
(2)i be the vertices of
ith copy of K2 attached with vi, for 1 ≤ i ≤ n. Define f : V (Pn ◦ K2) → {1, 2, 3, · · · , 4n} as
follows:
f(vi) = 4i− 2, for 1 ≤ i ≤ n,
f(u(1)i ) = 4i− 3, for 1 ≤ i ≤ n,
f(u(2)i ) = 4i, for 1 ≤ i ≤ n.
Then the induced edge labeling f∗ is obtained as follows:
f∗(vivi+1) = 4i, for 1 ≤ i ≤ n− 1,
f∗(u(1)i u
(2)i ) = 4i− 2, for 1 ≤ i ≤ n,
f∗(u(1)i v1) = 4i− 3, for 1 ≤ i ≤ n,
f∗(u(2)i vi) = 4i− 1, for 1 ≤ i ≤ n.
Hence f is an F -centroidal mean labeling of Pn ◦K2 for n ≥ 1. Thus the graph Pn ◦K2 is an
F -centroidal mean graph, for n ≥ 1. 2AAAAAAs s AAAAAAs s AAAAAAs s s s AAAAAAs s2 6 10 14 18
1 4 5 8 9 12 13 1617 20
1 3 5 7 9 11 13 15 17 19
4 8 12 16 ss s s s6 10 14 182
Figure 6 An F -centroidal mean labeling of P5 ◦K2
Theorem 2.7 The graph P ba is an F -centroidal mean graph, for a ≥ 2 and b ≤ 3.
Proof Let yi, xij1, xij2, . . . xiji, yi+1 be the vertices of the path P ji , where 1 ≤ i ≤ a−1 and
1 ≤ j ≤ b. Let V (P ba) = {yi : 1 ≤ i ≤ a} ∪
a−1⋃i=1
b⋃j=1
{xijk : 1 ≤ k ≤ i} and E(P ba) =
a−1⋃i=1
{yixij1 :
1 ≤ i ≤ b} ∪a−1⋃i=1
b⋃j=1
{xijkxij(k+1) : 1 ≤ k ≤ i− 1} ∪a−1⋃i=1
{xijiyi+1 : 1 ≤ j ≤ b} be the vertex set
and edge set of the graph P ba .
Case 1. b = 2.
Define f : V (P 2a ) → {1, 2, 3, · · · , (a− 1)(a+ 2) + 1} as follows:
f(y1) = 1,
f(yi) = (i− 1)(i+ 2) + 1, for 2 ≤ i ≤ a,
f(x1j1) = j + 1, for 1 ≤ j ≤ 2 and for 2 ≤ i ≤ a− 1,
F -Centroidal Mean Labeling of Graphs Obtained From Paths 129
f(xijk) = (i− 1)(i+ 2) + 2k + j − 1, for 1 ≤ k ≤ i and 1 ≤ j ≤ 2.
Then the induced edge labeling f∗ is obtained as follows:
f∗(y1x1j1) = j, for 1 ≤ j ≤ 2,
f∗(x1j1y2) = j + 2, for 1 ≤ j ≤ 2,
f∗(yixij1) = (i− 1)(i+ 2) + j, for 2 ≤ i ≤ a− 1 and 1 ≤ j ≤ 2,
f∗(xijkxij(k+1)) = (i− 1)(i+ 2) + j + 2k, for 2 ≤ i ≤ a− 1, 1 ≤ k ≤ i− 1 and 1 ≤ j ≤ 2,
f∗(xijiyi+1) = i(i+ 3) + j − 2, for 2 ≤ i ≤ a− 1 and 1 ≤ j ≤ 2.
Case 2. b = 3.
Define f : V (P 3a ) → {1, 2, 3, · · · , 3(a−1)(a+2)
2 + 1} as follows:
f(y1) = 1, f(y2) = 5, f(x111) = 2, f(x1j1) = 4j − 5, for 2 ≤ j ≤ 3,
f(yi) =3(i− 1)(i+ 2)
2+ 1, for 3 ≤ i ≤ a, f(x21k) =
4k + 5, k = 1,
4k + 4, k = 2,
f(x22k) =
7k + 6, k = 1,
7k − 4, k = 2,f(x23k) =
5k + 6, k = 1,
5k + 4, k = 2.
For 3 ≤ i ≤ a− 1,
f(xij1) =
3(i−1)(i+2)2 + j + 1, 1 ≤ j ≤ 2,
3(i−1)(i+2)2 + 2j, j = 3 and
f(xijk) =
3(i−1)(i+2)2 + 2j + 3k − 1, 1 ≤ j ≤ 2, 2 ≤ k ≤ i− 1,
and k is even,
3(i−1)(i+2)2 + 3k − 2, j = 3, 2 ≤ k ≤ i− 1
and k is even,
3(i−1)(i+2)2 + 2j + 3k − 3, 1 ≤ j ≤ 3, 2 ≤ k ≤ i− 1
and k is odd,3(i−1)(i+2)
2 + 3k − 1, j = 1, k = i and k is odd,3(i−1)(i+2)
2 + 3k + j − 1, j = 2, k = i and k is odd,3(i−1)(i+2)
2 + 3k + j, j = 3, k = i and k is odd,3(i−1)(i+2)
2 + 3k + j, 1 ≤ j ≤ 2, k = i,
and k is even,3(i−1)(i+2)
2 + 3k − 1, j = 3, k = i and k is even.
130 S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar
Then the induced edge labeling f∗ is obtained as follows:
f∗(y1x1j1) =
j, 1 ≤ j ≤ 2,
5, j = 3,
f∗(yixij1) =
3(i−1)(i+2)2 + j, j = 1 and 2 ≤ i ≤ a− 1,
3(i−1)(i+2)2 + j + 1, j = 2 and i = 2,
3(i−1)(i+2)2 + j − 1, j = 3 and i = 2,
3(i−1)(i+2)2 + j, j = 2, 3 and 3 ≤ i ≤ a− 1,
f∗(x1j1y2) =
3, j = 1,
2j, 2 ≤ j ≤ 3,f∗(x2j2y3) =
14, j = 1,
13, j = 2,
15, j = 3,
f∗(x2j1x2j2) = j + 9 for 1 ≤ j ≤ 3 and 3 ≤ i ≤ a− 1,
f∗(xijkxij(k+1)) =
3(i−1)(i+2)2 + 3k + 2(j − 1) + 1, 1 ≤ k ≤ i− 1,
and 1 ≤ j ≤ 2,3(i−1)(i+2)
2 + 3k + 2, 1 ≤ k ≤ i− 1,
and j = 3,
and f∗(xijiyi+1) =
3i(i+3)2 + j − 3, 1 ≤ j ≤ 3 and i is odd,
3i(i+3)2 + j − 2, 1 ≤ j ≤ 2 and i is even,
3i(i+3)2 − 2, j = 3 and i is even.
Hence f is an F -centroidal mean labeling of P ba for a ≥ 2 and b ≤ 3. Thus the graph P b
a is an
F -centroidal mean graph, for a ≥ 2 and b ≤ 3. 21
2
3
8
11
12
14
16
17
15
13
1 3
2 4
13 15
17
12
14 16
18
5
7
6
11
195
79
68
10
s s s s s sss9
s s s sssss
s ss1
2
7
3
9 12
11 14
13 10
17 22 23
21 19 27
18 24 259 11 13 17 21 24 262 4
1 3
5 6
710
14
8
12
15
1619 22
18
20 23
27
25
s s ss ss s 28s5 16
s ss ss s ss s ssFigure 7 An F -centroidal mean labeling of P 2
4 and P 34
F -Centroidal Mean Labeling of Graphs Obtained From Paths 131
Theorem 2.8 The middle graph M(Pn) of a path Pn is an F -centroidal mean graph.
Proof Let V (Pn) = {v1, v2, v3, · · · , vn} and E(Pn) = {ei = vivi+1 : 1 ≤ i ≤ n− 1} be the
vertex set and edge set of the path Pn. Then,
V (M(Pn)) = {v1, v2, v3, . . . , vn, e1, e2, e3, · · · , en−1},E(M(Pn)) = {viei, eivi+1 : 1 ≤ i ≤ n− 1} ∪ {eiei+1 : 1 ≤ i ≤ n− 2}.
Define f : V (M(Pn)) → {1, 2, 3, · · · , 3n− 3} as follows:
f(vi) =
1, for i = 1,
3i− 3, for 2 ≤ i ≤ n,
f(ei) = 3i− 1, for 1 ≤ i ≤ n− 1.
Then the induced edge labeling f∗ is obtained as follows:
f∗(viei) = 3i− 2, for 1 ≤ i ≤ n− 1,
f∗(eivi+1) = 3i− 1, for 1 ≤ i ≤ n− 1,
f∗(eiei+1) = 3i, for 1 ≤ i ≤ n− 2.
Hence f is an F -centroidal mean labeling of the graph M(Pn). Thus the graph M(Pn) is an
F -centroidal mean graph. 26
uuuu uu1 3 9 12
2 5 8 11
12 4 5
78
10 11u63 9
u uFigure 8 An F -centroidal mean labeling of M(P5)
Theorem 2.9 The total graph T (Pn) of a path Pn is an F -centroidal mean graph for n ≥ 1.
Proof Let V (Pn) = {v1, v2, v3, · · · , vn} and E(Pn) = {ei = vivi+1 : 1 ≤ i ≤ n− 1} be the
vertex set and edge set of the path Pn. Then V (T (Pn)) = {v1, v2, v3, · · · , vn, e1, e2, e3, · · · , en−1}and E(T (Pn)) = {vivi+1, eivi, eivi+1 : 1 ≤ i ≤ n− 1} ∪ {eiei+1 : 1 ≤ i ≤ n− 2}.
Define f : V (T (Pn)) → {1, 2, 3, . . . , 4(n− 1)} as follows:
f(v1) = 1,
f(vi) = 4i− 4, for 2 ≤ i ≤ n,
f(ei) = 4i− 2, for 1 ≤ i ≤ n− 1.
132 S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar
Then the induced edge labeling f∗ is obtained as follows:
f∗(vivi+1) = 4i− 2, for 1 ≤ i ≤ n− 1,
f∗(eiei+1) = 4i, for 1 ≤ i ≤ n− 2,
f∗(viei) = 4i− 3, for 1 ≤ i ≤ n− 1,
f∗(eivi+1) = 4i− 1, for 1 ≤ i ≤ n− 1.
Hence f is an F -centroidal mean labeling of the graph T (Pn). Thus the graph T (Pn) is an
F -centroidal mean graph. 21 4 12 16 208
2 6 10 14 18
2 6 10 14 18
84 12 16
1 5 9 13 17
3 7 11 15 19
t t ttt t t ttt tFigure 9 An F -centroidal mean labeling of T (P6)
Theorem 2.10 The square graph P 2n of the path Pn is an F -centroidal mean graph for n ≥ 1.
Proof Let v1, v2, v3, · · · , vn be the vertices of the path Pn.Define f : V(P 2
n
)→ {1, 2, 3, · · · , 2(n−
1)} as follows:
f(v1) = 1,
f(vi) = 2i− 2, for 2 ≤ i ≤ n.
Then the induced edge labeling f∗ is obtained as follows:
f∗(vivi+1) = 2i− 1, for 1 ≤ i ≤ n− 1,
f∗(vivi+2) = 2i, for 1 ≤ i ≤ n− 2.
Hence f is an F -centroidal mean labeling of the graph P 2n . Thus the graph P 2
n is an F -centroidal
mean graph. 21
2
4
6
8
1012
1 3 5 7 9 11
2 106
4 8t t t t tttFigure 10 An F -centroidal mean labeling of P 2
7
F -Centroidal Mean Labeling of Graphs Obtained From Paths 133
Theorem 2.11 The splitting graph S′(Pn) is an F -centroidal mean graph for n ≥ 2.
Proof Let v1, v2, · · · , vn be the vertices of the path Pn. Let v1, v2, · · · , vn, v′1, v
′2, · · · , v′n
be the vertices of the graph S′(Pn). Let V (S′(Pn)) = {vi, v′i : 1 ≤ i ≤ n} and E(S′(Pn)) =
{vivi+1, viv′i+1, v
′ivi+1 : 1 ≤ i ≤ n − 1} be the vertex set and edge set of the splitting graph
S′(Pn).
Case 1. n is odd.
Define f : V (S′(Pn)) → {1, 2, 3, · · · , 3n− 2} as follows:
f(vi) =
4i− 3, 1 ≤ i ≤ 2,
3, i = 3,
3i− 4, 4 ≤ i ≤ n and i is odd,
3i, 4 ≤ i ≤ n and i is even,
f(v′i) =
6, i = 1,
2, i = 2,
3i− 2, 3 ≤ i ≤ n.
Then the induced edge labeling f∗ is obtained as follows:
f∗(vivi+1) =
i+ 2, 1 ≤ i ≤ 2,
3i− 1, 3 ≤ i ≤ n− 1,
f∗(viv′i+1) =
5i− 4, 1 ≤ i ≤ 2,
3i− 2, 3 ≤ i ≤ n− 1 and i is odd,
3i, 3 ≤ i ≤ n− 1 and i is even,
f∗(v′ivi+1) =
5, i = 1,
2, i = 2,
3i, 3 ≤ i ≤ n− 1 and i is odd,
3i− 2, 3 ≤ i ≤ n− 1 and i is even.
Case 2. n is even.
Define f : V (S′(Pn)) → {1, 2, 3, · · · , 3n− 2} as follows:
f(vi) =
4 − i, 1 ≤ i ≤ 2,
3i− 1, 3 ≤ i ≤ n and i is odd,
3i− 3, 3 ≤ i ≤ n and i is even,
134 S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar
f(v′i) =
1, i = 1,
3i− 3, 2 ≤ i ≤ n and i is odd,
3i− 2, 2 ≤ i ≤ n and i is even.
Then the induced edge labeling f∗ is obtained as follows:
f∗(vivi+1) = 3i− 1, for 1 ≤ i ≤ n− 1,
f∗(viv′i+1) =
3i, 3 ≤ i ≤ n− 1 and i is odd,
3i− 2, 3 ≤ i ≤ n− 1 and i is even,
f(v′ivi+1) =
3i− 2, 1 ≤ i ≤ n− 1 and i is odd,
3i, 1 ≤ i ≤ n− 1 and i is even.
Hence f is an F -centroidal mean labeling of S′(Pn). Thus the splitting graph S′(Pn) is an
F -centroidal mean graph for n ≥ 2. 2
������ ������1 4 6 10 12 16 18
22
3 2 8 9 14 15 20 212 5 8 11 14 17 20
3 4 9 10 15 16 21
1 6 7 12 13 18 19
������1 5 3 12 11 18
6 2 7 1013
16 19
3 4 8 14
1
5 2 9 10 15 16
6 712 13 18
171711
s s sss ss s s s s ss s
s s s ss s s s s ssss s s s
Figure 11 An F -centroidal mean labeling of S′(P7) and S′(P8)
Theorem 2.12 The graph P (1, 2, · · · , n− 1) is an F -centroidal mean graph for n ≥ 2.
Proof Let v1, v2, · · · , vn be the vertices of the path Pn and let uij be the vertices of
the partition of K2,miwith cardinality mi, 1 ≤ i ≤ n − 1 and 1 ≤ j ≤ mi. Define f :
V (P (1, 2, · · · , n− 1)) → {1, 2, 3, · · · , n(n− 1) + 1} as follows:
f(vi) = i(i− 1) + 1, for 1 ≤ i ≤ n,
f(uij) = i(i− 1) + 2j, for 1 ≤ j ≤ i, and 1 ≤ j ≤ n− 1.
F -Centroidal Mean Labeling of Graphs Obtained From Paths 135
Then the induced edge labeling f∗ is obtained as follows:
f∗(viuij) = i(i− 1) + j, for 1 ≤ j ≤ i and 1 ≤ i ≤ n− 1,
f(uijvi+1) = i2 + j, for 1 ≤ j ≤ i and 1 ≤ i ≤ n− 1.
Hence f is an F -centroidal mean labeling of P (1, 2, . . . , n−1). Thus the graph P (1, 2, . . . , n−1)
is an F -centroidal mean graph for n ≥ 2. 2AAAA SSSSq q q qq1
2
3
4 8
14
20
18
7 13 21
1 2 3 75 10 13
17
6 12
96 12 15 19
20
16
14188 11
q qqq22
31
30
28
26
2421
22
26
27
3025
24 2916
23 28
4
q qq q qqqqq
BBBBBBqq10q
Figure 12 An F -centroidal mean labeling of P (1, 2, 3, 4, 5)
References
[1] S.Arockiaraj, A.Durai Baskar and A.Rajesh Kannan, F -root square mean labeling of
graphs obtained from paths, International Journal of Mathematical Combinatorics, Vol.2
(2017), 92–104.
[2] S.Arockiaraj, A.Durai Baskar and A.Rajesh Kannan, F -root square mean labeling of some
cycle related graphs, Journal of Engineering Technology, 6 (2017), 440–448.
[3] A.Durai Baskar and P.Manivannan, F -heronian mean labeling of graphs obtained from
paths, International Journal of Pure and Applied Mathematics, 117(5) (2017), 55–62.
[4] A.Durai Baskar, S.Arockiaraj and B.Rajendran, Geometric meanness of graphs obtained
from paths, Util. Math., 101(2016), 45–68.
[5] A.Durai Baskar and S.Arockiaraj, F -geometric mean graphs, International Journal of Ap-
plications and Applied Mathematics, 10(2) (2015), 937–952.
[6] A.Durai Baskar and S.Arockiaraj, F -harmonic mean graphs, International Journal of
Mathematical Archieve, 5(8) (2014), 92–108.
[7] J.A.Gallian, A Dynamic Survey of Graph Labeling, The Electronic Journal of Combina-
torics, 17(2017), #DS6.
[8] F.Harary, Graph Theory, Narosa Publishing House Reading, New Delhi, 1988.
International J.Math. Combin. Vol.4(2019), 136-148
Further Results on Analytic Odd Mean Labeling of Graphs
P.Jeyanthi1,R.Gomathi2 and Gee-Choon Lau3
1. Research Centre, Department of Mathematics
Govindammal Aditanar College for Women, Tiruchendur 628215, Tamil Nadu, India
2. Research Scholar, Reg.No.11825
Manonmaniam Sundaranar University, Abishekapatti, Tirunelveli-627 012, Tamil Nadu, India
3. Faculty of Computer and Mathematical SciencesUniversiti Tek nologi MARA (Segamat Campus), 85009, Johor, Malaysia
E-mail: [email protected], [email protected], [email protected]
Abstract: Let G = (V, E) be a graph with p vertices and q edges. A graph G is analytic
odd mean if there exist an injective function f : V → {0, 1, 3, 5 · · · , 2q − 1} with an induce
edge labeling f∗ : E → Z such that for each edge uv with f(u) < f(v),
f∗(uv) =
⌈f(v)2−(f(u)+1)2
2
⌉, if f(u) 6= 0
⌈f(v)2
2
⌉, if f(u) = 0
is injective. We say that f is an analytic odd mean labeling of G. In this paper we prove
that sun graph Sn, prism Dn, helm graph Hn, the graph Cn ◦P2, banana tree, bamboo tree,
perfect binary tree, the graph PCn, unicyclic graph, the caterpillar Pk(n0, n1, · · · , nk−1) and
spider graph are analytic odd mean graph.
Key Words: Mean labeling, analytic mean labeling, analytic odd mean labeling, Smaran-
dachely analytic odd mean labeling, analytic odd mean graph.
AMS(2010): 05C78.
§1. Introduction
Throughout this paper we consider only finite, simple and undirected graph G = (V,E) with
p vertices and q edges and notations not defined here are used in the sense of Harary [2].
A graph labeling is an assignment of integers to the vertices or edges or both, subject to
certain conditions. There are several types of labeling. An excellent survey of graph labeling
is available in [1]. The concept of analytic mean labeling was introduced in [7]. A graph
G is analytic mean graph if it admits a bijection f : V → {0, 1, 2, · · · , p− 1} such that the
induced edge labeling f∗ : E → Z given by f∗(uv) =⌈
f(u)2−f(v)2
2
⌉with f(u) > f(v) is
injective. Motivated by the results in [7], we introduced a new mean labeling called analytic
odd mean labeling in [3]. A graph G is an analytic odd mean if there exist an injective function
f : V → {0, 1, 3, 5 · · · , 2q − 1} with an induce edge labeling f∗ : E → Z such that for each edge
1Received July 24, 2019, Accepted December 10, 2019.
Further Results on Analytic Odd Mean Labeling of Graphs 137
uv with f(u) < f(v),
f∗(uv) =
⌈f(v)2−(f(u)+1)2
2
⌉, if f(u) 6= 0
⌈f(v)2
2
⌉, if f(u) = 0
is injective. We say such an f is an analytic odd mean labeling of G. Otherwise, a Smaran-
dachely analytic odd mean labeling of G if there exists an integer 0 ≤ k ≤ q holding with
|f−∗(k)| ≥ 2, where f−∗ is the inverse of f∗. We proved that cycle Cn, path Pn, n-bistar, comb
Pn ⊙K1, graph Ln ⊙K1, wheel graph Wn, flower graph Fln, some splitting graphs, multiple of
graphs, quadrilateral snake Q(n), double quadrilateral snake DQ(n), coconut tree, fire cracker
and some star graphs, splitting graph spl(G), Pn(1, 2, 3, · · ·, n), the complete bipartite graph
Km,n, the graph Ck⊙Kn, the square graph of Pn, Cn, Bn,n, H-graph and H⊙mK1 are analytic
odd mean graphs in [4], [5] and [6].
We use the following definitions in the subsequent section to prove the results.
Definition 1.1 A sun graph Sn is a cycle Cn with a pendent edge attached to each vertex of a
cycle Cn.
Definition 1.2 The prism Dn, n ≥ 3 is a cubic graph which can be represented as a Cartesian
product P2 × Cn of a path on two vertices with a cycle on n vertices.
Definition 1.3 A banana tree is a tree obtained from a family of stars by joining one end
vertex of each star to a new vertex.
Definition 1.4 A tree is called a spider if it has a center vertex c of degree R > 1 and all the
other vertex is either a leaf or with degree 2. Thus a spider is an amalgamation of k paths with
various lengths. If it has x1’s path of length a1, x2’s path of length a2,· · · . We denote the spider
graph by SP (a1x1 , a2
x2 , · · · , amxm) where a1 < a2 < · · · < am and x1 + x2 + · · · + xm = R.
Definition 1.5 A helm Hn, n ≥ 3 is obtained from the wheel graph Wn by adding a pendent
edge at each vertex on the wheel’s rim.
Definition 1.6 A bamboo tree (Pn ⊗ K1,ni)ki=1 is an one point union of Pn ⊗ K1,ni
where
1 ≤ i ≤ k.
Definition 1.7 A caterpillar Pk(n1, n2, · · · , nk) is a tree in which all the vertices are within
distance 1 of a central path Pk for k ≥ 1. When k ≥ 2, a caterpillar is obtained from a path
Pk = u1u2u3 · · ·uk attaching ni ≥ 0 pendent vertices vji (1 ≤ j ≤ ni) to each ui.
Definition 1.8 A perfect binary tree is a full binary tree in which all the leaves are at the same
level and in which every parent has two children.
§2. Main Results
In this section we prove that sun graph Sn, prism Dn, helm graph Hn, path union of n − 3
138 P.Jeyanthi, R.Gomathi and Gee-Choon Lau
copies of Cn,the graph Cn ◦ P2, banaba tree, bamboo tree, the graph PCn, unicyclic graph,
perfect binary tree, the caterpillar graph Pk(n0, n1, · · · , nk−1) and spider graph are analytic
odd mean graphs.
Theorem 2.1 For every positive integer n ≥ 3, the sun graph Sn is an analytic odd mean
graph.
Proof Let the vertex set and edge set of the sun graph be V (Sn) = {ui, vi : 1 ≤ i ≤ n}and E(Sn) = {uiui+1 : 1 ≤ i ≤ n− 1}⋃ {uivi : 1 ≤ i ≤ n}⋃ {unu1}Now |V (G)| = 2n = |E(G)|. We define an injective map f : V (Sn) → {0, 1, 3, 5, · · · , 4n− 1} by
f(ui) = 2i− 1 for 1 ≤ i ≤ n and f(vi) = 2n+ 2i− 1 for 1 ≤ i ≤ n.
The induced edge labeling f∗ is defined as follows:
f∗(uiui+1) = 2i+ 1 for 1 ≤ i ≤ n− 1,
f∗(u1un) = 2n2 − 2n− 1,
f∗(uivi) = 2n(n− 1) + 2i(2n− 1) + 1 for 1 ≤ i ≤ n.
We observe that the edge labels of uiui+1 are 3, 5, · · · 2n − 1 as i increases and the edge
labels of uivi are increased by 4n− 2 as i increases from 1 to n. Hence all the edge labels are
distinct and odd . Hence Sn admits an analytic odd mean labeling. 2An analytic odd mean labeling of S8 is shown in Figure 1.
b b
b
b
b
b
b b
b
b
b
b
b
b
b
b1
3
5
79
11
13
15
17
3
21
2325
27
29
31
5
7
9
11
13
15
111
143
173
203
233263
293
323
353
19
Figure 1
Theorem 2.2 For every positive n ≥ 3, the prism Dn is an analytic odd mean graph.
Proof Let the vertex set and edge set of the prism be V (Dn) = {ui, vi : 1 ≤ i ≤ n} and
E(Dn) = {uiui+1, vivi+1 : 1 ≤ i ≤ n−1}⋃{uivi : 1 ≤ i ≤ n}⋃ {unv1, vnv1}. Now |V (Dn)| =
2n and |E(Dn)| = 3n. We define an injective map f : V (Dn) → {0, 1, 3, 5, · · · , 6n − 1} by
f(ui) = 2i− 1 for 1 ≤ i ≤ n and f(vi) = 2n+ 2i− 1 for 1 ≤ i ≤ n.
Further Results on Analytic Odd Mean Labeling of Graphs 139
The induced edge labeling f∗ is defined as follows:
f∗(uiui+1) = 2i+ 1 for 1 ≤ i ≤ n− 1,
f∗(u1un) = 2n2 − 2n− 1,
f∗(uivi) = 2n(n− 1) + 2i(2n− 1) + 1 for 1 ≤ i ≤ n,
f∗(vivi+1) = 2n+ 2i+ 1 for 1 ≤ i ≤ n− 1,
f∗(v1vn) = 6n2 − 8n− 1.
We observe that the edge labels of uiui+1 and vivi+1 are 3, 5, . . . , 2n− 1 and 2n+ 3, 2n+
5, · · · , 2n+ 2n− 1 = 4n− 1 respectively as i increases and the edge labels of uivi are increased
by 4n − 2 as i increases from 1 to n. So all the edge labels are distinct and odd. Hence Dn
admits an analytic odd mean labeling. 2An analytic odd mean labeling of D8 is shown in Figure 2.
b b
b
b
b
b
b b
b
b
b
b
b
b
b
b1
3
5
79
11
13
15
17
3
21
2325
27
29
31
5
7
9
11
13
15
111
143
173
203
233263
293
323
353
19
19
21
23
25
27
29
31
319
Figure 2
Theorem 2.3 The helm graph Hn, n ≥ 3 is an analytic odd mean graph.
Proof Let V (Hn) = {v, vi, ui : 1 ≤ i ≤ n} and E(Hn) = {vvi, viui : 1 ≤ i ≤ n} ∪{vivi+1 : 1 ≤ i ≤ n − 1} ∪ {vnv1}. Now |V (G)| = 2n + 1 and |E(G)| = 3n. We define an
injective map f : V (G) → {0, 1, 3, 5, · · · , 6n− 1} by f(v) = 0, f(vi) = 4i− 3 for 1 ≤ i ≤ n and
f(ui) = 4n+ 2i− 1 for 1 ≤ i ≤ n.
The induced edge labeling f∗ is defined as follows:
f∗(vvi) = 8i2 − 12i+ 5 for 1 ≤ i ≤ n,
f∗(vivi+1) = 12i− 1 for 1 ≤ i ≤ n− 1,
f∗(vnv1) = 8n2 − 12n+ 3,
f∗(viui) = 4n(2n− 1) − 6i(i− 1) + 8ni− 1 for 1 ≤ i ≤ n.
140 P.Jeyanthi, R.Gomathi and Gee-Choon Lau
We observe that when i increases, the difference of edges viui are decreased by 12. Clearly
all the edge labels are odd and distinct. Hence Hn admits an analytic odd mean labeling. 2An analytic odd mean labeling of H10 is shown in Figure 3.
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b
b b
b
b
b
v
v9
v8
v7
v6
v5
v4
v3
v2
v1
u10
u9
u8
u7
u6
u5
u4
u3
u2
u1
v10
0
57
53
51
49
47
45
43
41
37
33
29
2521
17
13
9
5
1
55
59
Figure 3
Theorem 2.4 The graph PCn(n ≥ 4 and n is even) is obtained from Cn = v1v2 · · · vnv1 by
adding the chords vi and vn−i+2 for 2 ≤ i ≤ l where l = n/2. Then the graph PCn is an
analytic odd mean graph.
Proof Let G = PCn. Let the vertex set and edge set of G be V (G) = {vi, : 1 ≤ i ≤ n}and E(G) = {vivi+1 : 1 ≤ i ≤ n − 1} ∪ {vivn−i+2 : 2 ≤ i ≤ l} ∪ {vnv1}. Then there are
n vertices and 3n/2 − 1 edges. We define an injective map f : V (G) = {0, 1, 3, · · ·3n− 3} by
f(vi) = 2i− 1 for 1 ≤ i ≤ n.
The induced edge labeling f∗ is defined as follows:
f∗(vivi+1) = 2i+ 1 for 1 ≤ i ≤ n,
f∗(vnv1) = 2n2 − 2n− 1,
f∗(vivn−i+2) = 2n(n+ 3) − 2i(2n+ 3) + 5 for 2 ≤ i ≤ l.
It can be easily verified that f is an analytic odd mean labeling and hence PCn is an
analytic odd mean graph. 2An analytic odd mean labeling of PC10 is shown in Figure 4.
Further Results on Analytic Odd Mean Labeling of Graphs 141
b
b
b
b
b
b
b
b
b
b
1
3
5
7
9
11
13
15
17
19 173
127
81
35
179 3
5
7
9
1113
15
17
19
Figure 4
Theorem 2.5 Let G be a unicycle graph with a cycle Ck = a1a2 · · · aka1 such that the vertex
ai is attached to ni pendent vertices. Then the unicycle graph admits an analytic odd mean
labeling.
Proof Let V (G) = {ai, aji : 1 ≤ i ≤ k and 1 ≤ j ≤ ni} and E(G) = {aiai+1 : 1 ≤
i ≤ k − 1} ∪ {aiaji : 1 ≤ i ≤ k and 1 ≤ j ≤ ni} ∪ {aka1}. We assume n0 = 0. Hence
|V | = n1 + n2 + · · · + nk + k = |E| .We define an injective map on the vertex set by f(ai) = 2i − 1 for 1 ≤ i ≤ k and
f(aji ) = 2k − 1 + 2
∑i−1r=0 nr + 2j for 1 ≤ i ≤ k and 1 ≤ j ≤ ni.
The induced edge labeling f∗ is defined as follows:
f∗(aiai+1) = 2i+ 1 for 1 ≤ i ≤ k − 1,
f∗(aka1) = 2k2 − 2k − 1,
f∗(aiaji ) = 2(k2 + j2 − i2) + 2(
∑i−1r=0 nr)
2 + 2(2k+ 2j − 1)∑j−1
i=1 ni + 4kj − 2(k+ j) + 1 for
1 ≤ i ≤ k and 1 ≤ j ≤ ni.
Clearly the edge labels are odd and distinct. Hence the unicycle graph admits an analytic
odd mean labeling. 2An analytic odd mean labeling of the unicycle graph with k = 6 is shown in Figure 5.
b
b b
b b
b
b b
b
bb
b
b
b
b
b
b
b
b
b
b
b
b
1
43
41
39
35
37
33
11
9
7
5
31 29 27 25
21
23
19
17
13 15
45
3
3
5
79
11
83 111
137
163
203
333281
247563
495
449389
853
769689
613
941
Figure 5
59
142 P.Jeyanthi, R.Gomathi and Gee-Choon Lau
Theorem 2.6 The caterpillar Pk(n0, n1, · · · , nk−1) is an analytic odd mean graph for k ≥2, ni ≥ 0.
Proof Let V (G) = {vi, vji : 0 ≤ i ≤ k − 1 and 1 ≤ j ≤ ni} and E(G) = {vi−1vi : 1 ≤
i ≤ k − 1} ∪ {vivji : 0 ≤ i ≤ k − 1 and 1 ≤ j ≤ ni}. Hence |V | = n0 + n1 + . . . + nk + k and
|E| = n0 + n1 + · · · + nk + k − 1. We define an injective map on the vertex set by
f(v0) = 0, f(vi) = 2i− 1 for 1 ≤ i ≤ k − 1,
f(vj0) = 2k + 2j − 3 for 1 ≤ j ≤ ni,
f(vji ) = 2k − 3 + 2
∑i−1r=0 nr + 2j for 1 ≤ i ≤ k − 1 and 1 ≤ j ≤ ni.
The induced edge labeling f∗ is defined as follows:
f∗(vi−1vi) = 2i− 1 for 1 ≤ i ≤ k − 1,
f∗(v0vj0) = 2k(k − 3) + 2j(j − 3) + 4kj + 5 for 1 ≤ j ≤ ni,
f∗(vivji ) =
[(2k − 3 + 2
∑i−1r=0 nr + 2j − 2i
)(2k − 3 + 2
∑i−1r=0 nr + 2j + 2i
)+ 1]÷ 2 for
1 ≤ i ≤ k − 1 and 1 ≤ j ≤ ni.
Clearly the edge labels are odd and distinct. Hence Pk(n0, n1, ..., nk−1) admits an analytic
odd mean labeling. 2An analytic odd mean labeling of P5(3, 2, 1, 5, 4) is shown in Figure 6.
bb b b b
b
b
b b b b b
b bb
b b
b b
b
3119171513
119
75310
332927
2523
21 3537
Figure 6
Theorem 2.7 The graph Cn ◦ P2 is an analytic odd mean graph.
Proof Let G be the graph Cn ◦ P2. Let V (G) = {vi, vi,1, vi,2 : 1 ≤ i ≤ n} and E(G) =
{vivi+1 : 1 ≤ i ≤ n− 1} ∪ {vnv1} ∪ {vivi,1, vivi,2, vi,1vi,2 : 1 ≤ i ≤ n}. Now |V (G)| = 3n and
|E(G)| = 4n.
We define an injective map f : V (G) → {0, 1, 3, 5, · · · , 8n− 1} by
f(vi) = 2i− 1 for 1 ≤ i ≤ n,
f(vi,1) = 4i+ 2n− 3 for 1 ≤ i ≤ n,
f(vi,2) = 4i+ 2n− 1 for 1 ≤ i ≤ n.
The induced edge labeling f∗ is defined as follows:
f∗(vivi+1) = 2i+ 1 for 1 ≤ i ≤ n− 1,
f∗(vnv1) = 2n2 − 2n+ 1,
f∗(vivi,1) = 6i(i− 2) + 2n(n− 3) + 8ni+ 5 for 1 ≤ i ≤ n,
f∗(vivi,2) = 2i(3i− 2) + 2n(n− 1) + 8ni+ 1 for 1 ≤ i ≤ n,
f∗(vi,1vi,2) = 4i+ 2n− 1 for 1 ≤ i ≤ n.
Further Results on Analytic Odd Mean Labeling of Graphs 143
Clearly the edge labels are odd and distinct. Hence Cn ◦ P2 admits an analytic odd mean
labeling. 2An analytic odd mean labeling of C5 ◦ P2 is shown in Figure 7.
b b
b b
b
b
b
b
b
bb
b
b b
b
21
17
13
29
27
25
23 21
19
17
15
1311
39
9
7
5
3
9
7 5
3
1
203
163
25
233
281
315
37129
59 83
105
137
Figure 7
Theorem 2.8 Let K1,n1 ,K1,n2 , · · · ,K1,nkbe a family of stars with vertex sets V (K1,nj
) ={aj , a
1j , a
2j , · · · , a
nj
j
}, and deg(aj) = nj , 1 ≤ j ≤ k. Let BT (n1, n2, · · · , nk) be a banana tree
obtained by adding a new vertex a and joning it to a11, a
12, a
13, · · · , a1
k. Then BT (n1, n2, · · · , nk)
admits an analytic mean labeling where nj is any positive integer.
Proof Let the vertex set and edge set be V ={a, aj, a
rj : 1 ≤ j ≤ k, 1 ≤ r ≤ nj
}and E ={
aa1j : 1 ≤ j ≤ k
}∪{aja
rj : 1 ≤ j ≤ k, 1 ≤ r ≤ nj
}respectively. We assume n0 = 0. Hence
|V | = n1 + n2 + · · · + nk + k + 1 and |E| = n1 + n2 + · · · + nk + k.
We define an injective function f on the vertex set of banana tree as follows :
f(a) = 0, f(aj) = 2j − 1 for 1 ≤ j ≤ k,
f(arj) = 2k − 1 + 2
∑j−1i=0 ni + 2r for 1 ≤ r ≤ nj , 1 ≤ j ≤ k.
The induced edge labeling f∗ is defined as follows:
f∗(ajarj) = 2(k2 + r2 − j2) + 2(
j−1∑
i=0
ni + j)(2k − 1 +
j−1∑
i=0
ni) + 2j
j−1∑
i=0
ni − 2k + 1
for 1 ≤ r ≤ nj , 1 ≤ j ≤ k and
f∗(aa1j) =
[(2k − 1 + 2
∑j−1i=0 ni + 2
)2
+ 1
]
2
= 2k(k + 1) + 2
j−1∑
i=0
ni(2k + 1) + 2(
j−1∑
i=0
ni)2 + 1
144 P.Jeyanthi, R.Gomathi and Gee-Choon Lau
for 1 ≤ j ≤ k.
Clearly the edge labels are odd and distinct. Therefore f is an analytic odd mean labeling
and hence the banana tree is an analytic odd mean graph. 2An analytic odd mean labeling of BT (3, 5, 6, 4) is shown in Figure 8.
b
b b
b b
b
bb
b
b b
b
b
b
b
b
b
b
b
b
b
b
b
23
21
1917
13
9
11
7
53
1
0
2931
33
35
37
2515
105
43
41
39
27
257
213
173137
8359
685
313113
41
39
893
809
729653
595
527
463403
347
295
Figure 8
Theorem 2.9 Let K1,n1 ,K1,n2 , · · · ,K1,nk,K1,k be a family of stars with vertex sets V (K1,nj
) ={vr
j : 0 ≤ r ≤ nj
}and deg(v0
j ) = nj , for 1 ≤ j ≤ k, V (K1,k) = {v, v1, v2, · · · , vk}. The bam-
boo tree (P2 ⊗ K1,nj) for j = 1, 2, . . . , k is obtained by joining the vertex v1, v2, . . . , vk with
v11 , v
12 , · · · , v1
k respectively. Clearly the number of vertices and edges of the bamboo tree are
n1 + n2 + · · · + nk + k + 1 and n1 + n2 + . . . + nk + k respectively. Then the bamboo tree
(P2 ⊗K1,nj) for j = 1, 2, · · · , k is an analytic odd mean graph.
Proof Let the vertex set and edge set be V ={v, vj , v
rj : 1 ≤ j ≤ k, 0 ≤ r ≤ nj
}and
E = {vvj : 1 ≤ j ≤ k}∪{vjv
1j : 1 ≤ j ≤ k
}∪{v0
j vrj : 1 ≤ j ≤ k, 1 ≤ r ≤ nj
}respectively. We
assume n0 = 0. Hence |V | = n1 + n2 + . . .+ nk + 2k + 1 and |E| = n1 + n2 + · · · + nk + 2k.
We define an injective function f on the vertex set of bamboo tree as follows:
f(v) = 0, f(vj) = 2j − 1 for 1 ≤ j ≤ k,
f(vrj ) = 4k − 1 + 2
∑j−1i=0 ni + 2r for 1 ≤ r ≤ nj , 1 ≤ j ≤ k,
f(v0j ) = 2k + 2j − 1 for 1 ≤ j ≤ k.
The induced edge labeling f∗ is defined as follows:
f∗(vvj) = 2j2 − 2j + 1 for 1 ≤ j ≤ k,
f∗(vjv1j ) =
[(4k + 1 + 2
∑j−1i=0 ni
)2
+ 1
]÷ 2 − 2j2 for 1 ≤ j ≤ k,
f∗(v0j v
rj ) =
[(4k − 1 + 2
∑j−1i=0 ni + 2r
)2
+ 1
]÷ 2 − 2(k2 + j2) − 4kj for 1 ≤ r ≤ nj , 1 ≤
j ≤ k.
Clearly the edge labels are odd and distinct. Therefore f is an analytic odd mean labeling
and hence the bamboo tree is an analytic odd mean graph. 2
Further Results on Analytic Odd Mean Labeling of Graphs 145
An analytic odd mean labeling of (P2 ⊗K1,nj) for j = 1, 2, 3, 4 with n1 = 5, n2 = 4, n3 =
3, n4 = 7 is shown in Figure 9.
bb
b
b
b
b
b b
b b b b
b
b
b
b
b
b
b
b
bb
b b
b
b
b
b
47
33
31
2925
23
21
19
9
41352717
7531
0
49
5153
151311
3937
43
45
v64v5
1
v41
v31
v21
v02v0
1
v14
v13v1
2v11
v4v3v2v1
v
v54
v74
v04
v33v2
3
v03
v32
v42
v22
v24
v34
v44
Figure 9
v
Theorem 2.10 The perfect binary tree T of order p is an analytic odd mean graph.
Proof Let V (T ) = {vi : 1 ≤ i ≤ p} and E(T ) = {viv2i, viv2i+1 : 1 ≤ i ≤ q/2}. Hence
|V | = p and |E| = p − 1. We define an injective map f : V (G) = {0, 1, 3, · · · , 2p− 3} by
f(v1) = 0 and f(vi) = 2i− 3 for 2 ≤ i ≤ p. The induced edge labeling f∗ is defined as follows:
f∗(v1v2) = 1,
f∗(v1v3) = 5,
f∗(viv2i) = 6i2 − 8i+ 3 for 2 ≤ i ≤ q/2,
f∗(viv2i+1) = 6i2 − 1 for 2 ≤ i ≤ q/2.
We observe that the difference of edge labels of viv2i and viv2i+1 is 8i − 4 as i increases
from 1 to q2 . Therefore the edge labels are odd and distinct. Hence the binary tree admits an
analytic odd mean labeling. 2An analytic odd mean labeling of T of order 15 is shown in Figure 10.
b
bb
b bb b
b b b b b b b b
0
19171513
11975
31
27252321
Figure 10
53
293241215171113 1499567
332311
1 5
146 P.Jeyanthi, R.Gomathi and Gee-Choon Lau
Theorem 2.11 The spider graph SP (1n, km), n ≥ 1, and k,m ≥ 2 is an analytic odd mean
graph.
Proof Let V (SP (1n, km)) = {v} ∪ {wa, : 1 ≤ a ≤ n} ∪ {vji : 1 ≤ i ≤ k and 1 ≤ j ≤ m}
and E(SP (1n, km)) = {vvj1, v
ji v
ji+1 : 1 ≤ i ≤ k − 1 and 1 ≤ j ≤ m} ∪ {vwa : 1 ≤ a ≤ n}.
We define an injective map f : V (SP (1n, km)) → {0, 1, 3, 5, · · · , 2(km + n) − 1} by f(v) =
0, f(vji ) = (j − 1)2k+ 2i− 1 for 1 ≤ i ≤ k, 1 ≤ j ≤ m and f(wa) = 2mk+ 2a− 1 for 1 ≤ a ≤ n.
The induced edge labeling f∗ is defined as follows:
f∗(vvj1) = 2k(j − 1)(k(j − 1) + 1) + 1 for 1 ≤ j ≤ m,
f∗(vji v
ji+1) = 2k(j − 1) + 2i+ 1 for 1 ≤ i ≤ k − 1 and 1 ≤ j ≤ m,
f∗(vwa) = 2mk(mk − 1) + 2a(a− 1) + 4mak + 1 for 1 ≤ a ≤ n.
It can be verified that the edge labels are odd and distinct. Hence SP (1n, km) is an analytic
odd mean graph. 2An analytic odd mean labeling of SP (1n, 65) is shown in Figure 11.
b
b
b
b
b b b b b b
b b bb b
b b b bb
b b b b b
b bb b b
b
b
b
b
b
b
b
51
29 31 33 35
232119171513
1197531
0
49
474543413937
25 27
63
61
57 595553
2mk + 2n− 1
Figure 11
Theorem 2.12 The spider graph SP (2n, km), n ≥ 1, k,m ≥ 2 is an analytic odd mean graph
if (a) k is even and m is any integer and (b) k is odd and m is even.
Proof Let V (SP (2n, km)) = {v} ∪ {wa,1, wa,2 : 1 ≤ a ≤ n} ∪ {vji : 1 ≤ i ≤ k and 1 ≤ j ≤
m} and E(SP (2n, km)) = {vvj1, v
ji v
ji+1 : 1 ≤ i ≤ k ans 1 ≤ j ≤ m} ∪ {vwa,1, wa,1wa,2 : 1 ≤
a ≤ n}. We define an injective map f : V (SP (1n, km)) → {0, 1, 3, 5, · · · , 2km + 4n − 1} by
f(v) = 0, f(vji ) = (j − 1)2k + 2i − 1 for 1 ≤ i ≤ k, 1 ≤ j ≤ m; f(wa,1) = 2mk + 4a − 3 for
1 ≤ a ≤ n and f(wa,2) = 2mk + 4a− 1 for 1 ≤ a ≤ n.
The induced edge labeling f∗ is defined as follows:
f∗(vvj1) = 2k(j − 1)(k(j − 1) + 1) + 1 for 1 ≤ j ≤ m,
f∗(vji v
ji+1) = 2k(j − 1) + 2i+ 1 for 1 ≤ i ≤ k − 1 and 1 ≤ j ≤ m,
f∗(vwa,1) = 2mk(mk − 3) + 4a(2a− 3) + 8mak + 5 for 1 ≤ a ≤ n,
f∗(wa,1wa,2) = 2mk + 4a− 1 for 1 ≤ a ≤ n.
Further Results on Analytic Odd Mean Labeling of Graphs 147
It can be verified that the edge labels are odd and distinct. Hence SP (2n, km) is an analytic
odd mean graph. 2An analytic odd mean labeling of SP (2n, 54) is shown in Figure 12.
b b b b b
bb b b b
b b b b b
b b b bb
b
b b
b b
b b
b
b
b
0
1 3 5 7 9
11 13 15 17 19
21 23 25 27 29
31 33 35 37 39
41 43
45 47
2mk + 4n− 12mk + 4n− 3
Figure 12
Theorem 2.13 The spider graph SP (1s, 2n, km), n, s ≥ 1, k,m ≥ 2 is an analytic odd mean
graph if (a) k is even and m is any integer and (b) k is odd and m is even.
Proof Let V (SP (1s, 2n, km)) = {v} ∪ {wa,1, wa,2 : 1 ≤ a ≤ n} ∪ {vji : 1 ≤ i ≤ k
and 1 ≤ j ≤ m} ∪ {ur : 1 ≤ r ≤ s} and E(SP (1s, 2n, km)) = {vvj1, v
ji v
ji+1 : 1 ≤
i ≤ k and 1 ≤ j ≤ m} ∪ {vwa,1, wa,1wa,2 : 1 ≤ a ≤ n}. We define an injective map
f : V (SP (1n, km)) → {0, 1, 3, 5, · · · , 2km + 4n + 2s − 1} by f(ur) = 2mk + 4n + 2r − 1 for
1 ≤ r ≤ s and f(v), f(vji ), f(wa,1) and f(wa,2) are defined as in Theorem 2.12. Then the
induced edge labeling f∗(vvj1), f
∗(vji v
ji+1), f
∗(vwa,1) and f∗(wa,1wa,2) are as in Theorem 2.12
and
f∗(vur) =(2mk + 2r + 4n− 1)
2+ 1
2
for 1 ≤ r ≤ s.
It can be verified that the edge labels are odd and distinct. Hence SP (1s, 2n, km) is an
analytic odd mean graph. 2An analytic odd mean labeling of SP (1s, 2n, 34) is shown in Figure 13.
b b b
b b
b b
b
b
b
b
b
b
b
b
b
bb b
b b b
bb b
b0
1 3 5
7 9 11
13 15 17
19 21 23
25 27
2mk + 4n− 3 2mk + 4n− 1
2mk + 4n+ 2s− 1
2mk + 4n+ 3
2mk + 4n+ 1
Figure 13
148 P.Jeyanthi, R.Gomathi and Gee-Choon Lau
References
[1] J.A. Gallian, A dynamic survey of graph labeling, The Electronic Journal of Combinatorics,
(2018), # DS6.
[2] F.Harary, Graph Theory, Addison-Wesley ,Reading, Massachusetts, 1972.
[3] P.Jeyanthi, R.Gomathi and Gee-Choon Lau, Analytic odd mean labeling of square and
H-graphs, International J.Math. Combin., Special Issue 1 (2018), 61-67.
[4] P.Jeyanthi,R.Gomathi and Gee-Choon Lau, Analytic odd mean labeling of some standard
graphs, Palestine Journal of Mathematics, Vol.8(1)(2019), 444-450.
[5] P.Jeyanthi,R.Gomathi and Gee-Choon Lau, Some results on analytic odd mean labeling of
graph, Bulletin of The International Mathematical Virtual Institute, Vol. 9(2019), 487-500.
[6] P.Jeyanthi,R.Gomathi and Gee-Choon Lau, Analytic odd mean labeling of some graphs,
Palestine Journal of Mathematics, Vol.8(2)(2019), 392-399.
[7] T.Tharmaraj and P.B.Sarasija, Analytic mean labelled graphs, International Journal of
Mathematical Archive, 5(6)(2014), 136-146.
Papers Published in IJMC, 2019
Vol.1,2019
1. Harmonic Flows Dynamics on Animals in Microscopic Level with Balance Recovery
Linfan MAO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01
2. Null Quaternionic Slant Helices in Minkowski Spaces
T.Kahraman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3. Unique Metro Domination Number of Circulant Graphs
B. Sooryanarayana and John Sherra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4. On Hemi-Slant Submanifold of Kenmotsu Manifold
Chhanda Patra, Barnali Laha and Arindam Bhattacharyya . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5. The Number of Rooted Nearly 2-Regular Loopless Planar Maps
Shude Long and Junliang Cai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
6. A Note on Common Fixed Points for (ψ, α, β)-Weakly Contractive Mappings in Gener-
alized Metric Space
Krishnadhan Sarkar and Kalishankar Tiwary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
7. Zk-Magic Labeling of Cycle of Graphs
P.Jeyanthi and K.Jeya Daisy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88
8. Topological Efficiency Index of Some Composite Graphs
K.Pattabiraman and T.Suganya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
9. Total Domination Stable Graphs
Shyama M.P. and Anil Kumar V. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
10. Centered Triangular Mean Graphs
P.Jeyanthi, R.Kalaiyarasi and D.Ramya. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .126
11. New Families of Odd Mean Graphs
G.Pooranam, R.Vasuki and S.Suganthi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .134
12. F -Root Square Mean Labeling of Some Graphs
R.Gopi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Vol.2,2019
1. Computing Zagreb Polynomials of Generalized xyz-Point-Line Transformation Graphs
T xyz(G) With z = −B. Basavanagoud and Anand P. Barangi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01
2. On (j,m) Symmetric Convex Harmonic Functions
Renuka Devi K, Hamid Shamsan and S. Latha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3. On M-Projective Curvature Tensor of a (LCS)n-Manifold
Divyashree G. and Venkatesha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4. D-homothetic Deformations of Lorentzian Para-Sasakian Manifold
150 International Journal of Mathematical Combinatorics
Barnali Laha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5. Position Vectors of the Curves in Affine 3-Space According to Special Affine Frames
Yılmaz TUNCER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6. A Generalization of Some Integral Inequalities for Multiplicatively P-Functions
Huriye Kadakal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
7. C-Geometric Mean Labeling of Some Cycle Related Graphs
A.Durai Baskar and S.Arockiaraj . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
8. Neighbourhood V4-Magic Labeling of Some Shadow Graphs
Vineesh K.P. and Anil Kumar V. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
9. Open Packing Number of Triangular Snakes
S.K.Vaidya and A.D.Parmar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
10. Second Status Connectivity Indices and its Coindices of Composite Graphs
K.Pattabiraman and A.Santhakumar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
11. A Note on Detour Radial Signed Graphs
K. V. Madhusudhan and S. Vijay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
12. A Note on 3-Remainder Cordial Labeling Graphs
R.Ponraj, K.Annathurai and R.Kala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Vol.3,2019
1. On the Involute D-scroll in Euclidean 3-Space
Seyda Kılıcoglu and Suleyman SENYURT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01
2. On Prime Graph PG2(R) of a Ring
Sandeep S.Joshi and Kishor F.Pawar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
3. Non-Holonomic Frame for a Finsler Space with Some Special (α, β)-Metric
C.K.Yadav and M.K.Gupta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4. R1 Concepts in Fuzzy Bitopological Spaces with Quasi-Coincidence Sense
Ruhul Amin and Sahadat Hossain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5. Separations Axioms (T1) on Fuzzy Bitopological Spaces in Quasi-Coincidence Sense
Saikh Shahjahan Miah, M. R. Amin and Md.Fazlul Hoque . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6. On Radicals for Ternary Semirings
Swapnil P. Wani and Kishor F. Pawar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
7. Computation of Misbalance Type Degree Indices of Certain Classes of Derived-Regular
Graph
V. Lokesha, K. Zeba Yasmeen, B. Chaluvaraju and T. Deepika . . . . . . . . . . . . . . . . . . . . . . . . 61
8. Reciprocal Transmission Hosoya Polynomial of Graphs
Harishchandra S. Ramane and Saroja Y. Talwar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
9. Generalized Common Neighbor Polynomial of Graphs
Shikhi M. and Anil Kumar V.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .80
10. On Status Coindex Distance Sum and Status Connectivity Coindices of Graphs
Kishori P. Narayankar, Dickson Selvan and Afework T. Kahsay . . . . . . . . . . . . . . . . . . . . . . . 90
11. Some Results on 4-Total Prime Cordial Graphs
R.Ponraj, J.Maruthamani and R.Kala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Papers Published In IJMC, 2019 151
12. Super F -Centroidal Mean Graphs
S.Arockiaraj, A.Rajesh Kannan and A.Durai Baskar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Vol.4,2019
1. Graphs, Networks and Natural Reality – from Intuitive Abstracting to Theory
Linfan MAO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01
2. Some Subclasses of Meromorphic with p-Valent q-Spirallike Functions
Read.S.A.Qahtan, Hamid Shamsan and S.Latha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3. On the Complexity of Some Classes of Circulant Graphs and Chebyshev Polynomials
S. N. Daoud and Muhammad Kamran Siddiqui . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4. Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach Spaces
G.S.Saluja . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
5. Algebraic Properties of the Path Complexes of Cycles
Seyed Mohammad Ajdani . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6. Some Classes of Analytic Functions with q-Calculus
Hamid Shamsan, Fuad S. M. Alsarari and S. Latha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
7. On Obic Algebras
Ilojide Emmanuel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
8. Prime BI-Idels of Po-Γ-Groupoids
J.Catherine Grace John and B. Elavarasan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
9. 4-Total Prime Cordiality of Certain Subdivided Graphs
R. Ponraj, J. Maruthamani and R. Kala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
10. The Signed Product Cordial for Corona of Paths and Fourth Power of Paths
S. Nada, A. Elrokh, A. Elrayes and A. Rabie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102
11. On the Eccentric Sequence of Composite Graphs
Yaser Alizadeh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
12. F -Centroidal Mean Labeling of Graphs Obtained From Paths
S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
13. Further Results on Analytic Odd Mean Labeling of Graphs
P.Jeyanthi, R.Gomathi and Gee-Choon Lau. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Famous Words
The reality of a thing T is the behavior with motivation of an abstracted complex network
in the microcosmic level. Certainly, there are more microcosmic observing datum on the units,
cells or microcosmic particles of matters by scientific instruments. A microcosmic science is such
a science established on the microcosmic datum of matters, including theory and experimental
subjects, which must be established over 1-dimensional skeleton or in other words, topological
graphs. Could we establish such a mathematics over topological graphs for microcosmic science?
The answer is positive inspired by the traditional Chinese medicine, i.e., 12 meridians theory.
(Extracted from the paper: Science’s Dilemma - a Review on Science with Applications,
Progress in Physics, Vol.15, 2(2019), 78-85.)
By Dr.Linfan MAO, a Chinese mathematician, philosophical critic.
Author Information
Submission: Papers only in electronic form are considered for possible publication. Papers
prepared in formats, viz., .tex, .dvi, .pdf, or.ps may be submitted electronically to one member
of the Editorial Board for consideration in the International Journal of Mathematical
Combinatorics (ISSN 1937-1055). An effort is made to publish a paper duly recommended
by a referee within a period of 3 − 4 months. Articles received are immediately put the ref-
erees/members of the Editorial Board for their opinion who generally pass on the same in six
week’s time or less. In case of clear recommendation for publication, the paper is accommo-
dated in an issue to appear next. Each submitted paper is not returned, hence we advise the
authors to keep a copy of their submitted papers for further processing.
Abstract: Authors are requested to provide an abstract of not more than 250 words, lat-
est Mathematics Subject Classification of the American Mathematical Society, Keywords and
phrases. Statements of Lemmas, Propositions and Theorems should be set in italics and ref-
erences should be arranged in alphabetical order by the surname of the first author in the
following style:
Books
[4]Linfan Mao, Combinatorial Geometry with Applications to Field Theory, InfoQuest Press,
2009.
[12]W.S.Massey, Algebraic topology: an introduction, Springer-Verlag, New York 1977.
Research papers
[6]Linfan Mao, Mathematics on non-mathematics - A combinatorial contribution, International
J.Math. Combin., Vol.3(2014), 1-34.
[9]Kavita Srivastava, On singular H-closed extensions, Proc. Amer. Math. Soc. (to appear).
Figures: Figures should be drawn by TEXCAD in text directly, or as EPS file. In addition,
all figures and tables should be numbered and the appropriate space reserved in the text, with
the insertion point clearly indicated.
Copyright: It is assumed that the submitted manuscript has not been published and will
not be simultaneously submitted or published elsewhere. By submitting a manuscript, the
authors agree that the copyright for their articles is transferred to the publisher, if and when,
the paper is accepted for publication. The publisher cannot take the responsibility of any loss
of manuscript. Therefore, authors are requested to maintain a copy at their end.
Proofs: One set of galley proofs of a paper will be sent to the author submitting the paper,
unless requested otherwise, without the original manuscript, for corrections after the paper is
accepted for publication on the basis of the recommendation of referees. Corrections should be
restricted to typesetting errors. Authors are advised to check their proofs very carefully before
return.
December 2019
Contents
Graphs, Networks and Natural Reality –from Intuitive Abstracting to Theory
By Linfan MAO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01
Some Subclasses of Meromorphic with p-Valent q-Spirallike Functions
By Read.S.A.Qahtan, Hamid Shamsan and S.Latha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
On the Complexity of Some Classes of Circulant Graphs and Chebyshev
Polynomials By S. N. Daoud and Muhammad Kamran Siddiqui. . . . . . . . . . . . . . . . . . . . . . .29
Some Fixed Point Results for Multivalued Nonexpansive Mappings in Banach
Spaces By G.S.Saluja . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Algebraic Properties of the Path Complexes of Cycles
By Seyed Mohammad Ajdani . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Some Classes of Analytic Functions with q-Calculus
By Hamid Shamsan, Fuad S. M. Alsarari and S. Latha . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
On Obic Algebras By Ilojide Emmanuel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Prime BI-Ideals of Po-Γ-Groupoids
By J.Catherine Grace John and B. Elavarasan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4-Total Prime Cordiality of Certain Subdivided Graphs
By R. Ponraj, J. Maruthamani and R. Kala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
The Signed Product Cordial for Corona of Paths and Fourth Power of Paths
By S. Nada, A. Elrokh, A. Elrayes and A. Rabie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
On the Eccentric Sequence of Composite Graphs By Yaser Alizadeh . . . . . . . . . . . 112
F -Centroidal Mean Labeling of Graphs Obtained From Paths
By S. Arockiaraj, A. Rajesh Kannan and A. Durai Baskar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
Further Results on Analytic Odd Mean Labeling of Graphs
By P.Jeyanthi, R.Gomathi and Gee-Choon Lau . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Papers Published in IJMC, 2019 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
An International Journal on Mathematical Combinatorics